ROS index logo ROS Index
  1. Home
  2. Packages
  3. tensorrt_yolox
 
No version for distro humble showing github. Known supported distros are highlighted in the buttons above.
Package symbol

tensorrt_yolox package from autodrrt repo

autonomous_emergency_braking control_performance_analysis control_validator external_cmd_selector joy_controller lane_departure_checker mpc_lateral_controller obstacle_collision_checker operation_mode_transition_manager pid_longitudinal_controller predicted_path_checker pure_pursuit shift_decider trajectory_follower_base trajectory_follower_node vehicle_cmd_gate diagnostic_converter kinematic_evaluator localization_evaluator planning_evaluator ekf_localizer geo_pose_projector gyro_odometer ar_tag_based_localizer landmark_manager localization_error_monitor localization_gnss localization_util ndt_scan_matcher pose2twist pose_initializer pose_instability_detector stop_filter tree_structured_parzen_estimator twist2accel yabloc_common yabloc_image_processing yabloc_monitor yabloc_particle_filter yabloc_pose_initializer map_height_fitter map_loader map_projection_loader map_tf_generator lanelet2_map_preprocessor ros2_bevdet bytetrack cluster_merger compare_map_segmentation crosswalk_traffic_light_estimator detected_object_feature_remover detected_object_validation detection_by_tracker elevation_map_loader euclidean_cluster front_vehicle_velocity_estimator ground_segmentation heatmap_visualizer image_projection_based_fusion lidar_apollo_instance_segmentation lidar_apollo_segmentation_tvm lidar_apollo_segmentation_tvm_nodes lidar_centerpoint lidar_centerpoint_tvm map_based_prediction multi_object_tracker object_merger object_range_splitter object_velocity_splitter occupancy_grid_map_outlier_filter probabilistic_occupancy_grid_map radar_crossing_objects_noise_filter radar_fusion_to_detected_object radar_object_clustering radar_object_tracker radar_tracks_msgs_converter shape_estimation simple_object_merger tensorrt_classifier tensorrt_yolo tensorrt_yolox tracking_object_merger traffic_light_arbiter traffic_light_classifier traffic_light_fine_detector traffic_light_map_based_detector traffic_light_multi_camera_fusion traffic_light_occlusion_predictor traffic_light_ssd_fine_detector traffic_light_visualization behavior_path_avoidance_by_lane_change_module behavior_path_avoidance_module behavior_path_external_request_lane_change_module behavior_path_goal_planner_module behavior_path_lane_change_module behavior_path_planner behavior_path_planner_common behavior_path_side_shift_module behavior_path_start_planner_module behavior_velocity_blind_spot_module behavior_velocity_crosswalk_module behavior_velocity_detection_area_module behavior_velocity_intersection_module behavior_velocity_no_drivable_lane_module behavior_velocity_no_stopping_area_module behavior_velocity_occlusion_spot_module behavior_velocity_out_of_lane_module behavior_velocity_planner behavior_velocity_planner_common behavior_velocity_run_out_module behavior_velocity_speed_bump_module behavior_velocity_stop_line_module behavior_velocity_template_module behavior_velocity_traffic_light_module behavior_velocity_virtual_traffic_light_module behavior_velocity_walkway_module costmap_generator external_velocity_limit_selector freespace_planner freespace_planning_algorithms mission_planner motion_velocity_smoother objects_of_interest_marker_interface obstacle_avoidance_planner obstacle_cruise_planner obstacle_stop_planner obstacle_velocity_limiter path_smoother planning_debug_tools planning_test_utils planning_topic_converter planning_validator route_handler rtc_interface rtc_replayer bezier_sampler frenet_planner path_sampler sampler_common scenario_selector static_centerline_optimizer surround_obstacle_checker gnss_poser image_diagnostics image_transport_decompressor imu_corrector livox_tag_filter pointcloud_preprocessor radar_scan_to_pointcloud2 radar_static_pointcloud_filter radar_threshold_filter radar_tracks_noise_filter tier4_pcl_extensions vehicle_velocity_converter autoware_auto_msgs_adapter bluetooth_monitor component_state_monitor default_ad_api ad_api_adaptors ad_api_visualizers automatic_pose_initializer diagnostic_graph_aggregator dummy_diag_publisher dummy_infrastructure duplicated_node_checker emergency_handler mrm_comfortable_stop_operator mrm_emergency_stop_operator system_error_monitor system_monitor topic_state_monitor velodyne_monitor accel_brake_map_calibrator external_cmd_converter raw_vehicle_cmd_converter steer_offset_estimator vehicle_info_util launch loop_sim ssh_machine gpudirect-dds autoware_ad_api_specs autoware_adapi_v1_msgs autoware_adapi_version_msgs autoware_auto_common autoware_auto_geometry autoware_auto_control_msgs autoware_auto_geometry_msgs autoware_auto_mapping_msgs autoware_auto_msgs autoware_auto_perception_msgs autoware_auto_planning_msgs autoware_auto_system_msgs autoware_auto_vehicle_msgs autoware_auto_perception_rviz_plugin autoware_auto_tf2 autoware_cmake autoware_lint_common autoware_utils lanelet2_extension autoware_common_msgs autoware_control_msgs autoware_localization_msgs autoware_map_msgs autoware_perception_msgs autoware_planning_msgs autoware_sensing_msgs autoware_system_msgs autoware_vehicle_msgs autoware_point_types autoware_testing bag_time_manager_rviz_plugin component_interface_specs component_interface_tools component_interface_utils cuda_utils fake_test_node geography_utils global_parameter_loader glog_component goal_distance_calculator grid_map_utils interpolation kalman_filter motion_utils object_recognition_utils osqp_interface path_distance_calculator perception_utils polar_grid qp_interface rtc_manager_rviz_plugin signal_processing tensorrt_common tier4_adapi_rviz_plugin tier4_api_utils tier4_automatic_goal_rviz_plugin tier4_autoware_utils tier4_calibration_rviz_plugin tier4_camera_view_rviz_plugin tier4_control_rviz_plugin tier4_datetime_rviz_plugin tier4_debug_rviz_plugin tier4_debug_tools tier4_localization_rviz_plugin tier4_perception_rviz_plugin tier4_planning_rviz_plugin tier4_screen_capture_rviz_plugin tier4_simulated_clock_rviz_plugin tier4_state_rviz_plugin tier4_system_rviz_plugin tier4_target_object_type_rviz_plugin tier4_traffic_light_rviz_plugin tier4_vehicle_rviz_plugin time_utils simulator_compatibility_test traffic_light_recognition_marker_publisher traffic_light_utils tvm_utility dma_customer_msg dma_transfer eagleye_coordinate eagleye_navigation eagleye_msgs eagleye_rt eagleye_can_velocity_converter eagleye_fix2kml eagleye_geo_pose_converter eagleye_geo_pose_fusion eagleye_gnss_converter eagleye_tf llh_converter morai_msgs mussp ndt_omp orocos_kdl python_orocos_kdl pointcloud_to_laserscan rtklib_bridge rtklib_msgs autoware_external_api_msgs autoware_iv_external_api_adaptor autoware_iv_internal_api_adaptor awapi_awiv_adapter tier4_api_msgs tier4_auto_msgs_converter tier4_control_msgs tier4_debug_msgs tier4_external_api_msgs tier4_hmi_msgs tier4_localization_msgs tier4_map_msgs tier4_perception_msgs tier4_planning_msgs tier4_rtc_msgs tier4_simulation_msgs tier4_system_msgs tier4_v2x_msgs tier4_vehicle_msgs tier4_autoware_api_launch tier4_control_launch tier4_localization_launch tier4_map_launch tier4_perception_launch tier4_planning_launch tier4_sensing_launch tier4_simulator_launch tier4_system_launch tier4_vehicle_launch fastrtps cyclonedds lanelet2 lanelet2_core lanelet2_examples lanelet2_io lanelet2_maps lanelet2_matching lanelet2_projection lanelet2_python lanelet2_routing lanelet2_traffic_rules lanelet2_validation sophus angles behaviortree_cpp_v3 bond bond_core bondcpp bondpy smclib test_bond cudnn_cmake_module diagnostic_aggregator diagnostic_common_diagnostics diagnostic_updater diagnostics self_test filters geodesy geographic_info geographic_msgs grid_map grid_map_cmake_helpers grid_map_core grid_map_costmap_2d grid_map_cv grid_map_demos grid_map_filters grid_map_loader grid_map_msgs grid_map_octomap grid_map_pcl grid_map_ros grid_map_rviz_plugin grid_map_sdf grid_map_visualization mrt_cmake_modules nav2_amcl nav2_behavior_tree nav2_behaviors nav2_bringup nav2_bt_navigator nav2_collision_monitor nav2_common nav2_controller nav2_core nav2_costmap_2d costmap_queue dwb_core dwb_critics dwb_msgs dwb_plugins nav2_dwb_controller nav_2d_msgs nav_2d_utils nav2_lifecycle_manager nav2_map_server nav2_msgs nav2_navfn_planner nav2_planner nav2_regulated_pure_pursuit_controller nav2_rotation_shim_controller nav2_rviz_plugins nav2_simple_commander nav2_smac_planner nav2_smoother nav2_system_tests nav2_theta_star_planner nav2_util nav2_velocity_smoother nav2_voxel_grid nav2_waypoint_follower navigation2 dynamic_edt_3d octomap octovis octomap_msgs osqp_vendor pacmod3_msgs pcl_msgs pcl_conversions pcl_ros perception_pcl point_cloud_msg_wrapper radar_msgs can_msgs rqt_tf_tree tensorrt_cmake_module topic_tools topic_tools_interfaces tvm_vendor cv_bridge image_geometry opencv_tests vision_opencv xacro rviz2 rviz_assimp_vendor rviz_common rviz_default_plugins rviz_ogre_vendor rviz_rendering rviz_rendering_tests rviz_visual_testing_framework dummy_perception_publisher fault_injection simple_planning_simulator image_view v4l2_camera aion_interface carla_vehicle_can cgi430_can_driver cgi610_driver ARS408_driver common_sensor_launch kunyi_sensor_kit_description kunyi_sensor_kit_launch classformsg node_v2x data_format_dump data_preprocess_launch lidar_centerpoint_collect lidar_saver message_sync time_cal camera_calibration direct_visual_lidar_calibration multi_lidar_calibration

ROS Distro
github

API Docs
Browse Code

Package Summary

Tags No category tags.
Version 0.0.1
License Apache License 2.0
Build type AMENT_CMAKE
Use RECOMMENDED

Repository Summary

Description
Checkout URI https://github.com/ieiauto/autodrrt.git
VCS Type git
VCS Version main
Last Updated 2025-05-30
Dev Status UNKNOWN
Released UNRELEASED
Tags No category tags.
Contributing Help Wanted (-)
Good First Issues (-)
Pull Requests to Review (-)

Package Description

tensorrt library implementation for yolox

Additional Links

No additional links.

Maintainers

  • Daisuke Nishimatsu
  • Dan Umeda
  • Manato Hirabayashi

Authors

  • Daisuke Nishimatsu
src/autodrrt.application/perception/autoware.perception/tensorrt_yolox/README.md

tensorrt_yolox

Purpose

This package detects target objects e.g., cars, trucks, bicycles, and pedestrians on a image based on YOLOX model.

Inner-workings / Algorithms

Cite

Zheng Ge, Songtao Liu, Feng Wang, Zeming Li, Jian Sun, “YOLOX: Exceeding YOLO Series in 2021”, arXiv preprint arXiv:2107.08430, 2021 [ref]

Inputs / Outputs

Input

Name Type Description
in/image sensor_msgs/Image The input image

Output

Name Type Description
out/objects tier4_perception_msgs/DetectedObjectsWithFeature The detected objects with 2D bounding boxes
out/image sensor_msgs/Image The image with 2D bounding boxes for visualization

Parameters

Core Parameters

Name Type Default Value Description
score_threshold float 0.3 If the objectness score is less than this value, the object is ignored in yolox layer.
nms_threshold float 0.7 The IoU threshold for NMS method

NOTE: These two parameters are only valid for “plain” model (described later).

Node Parameters

Name Type Default Value Description
model_path string ”” The onnx file name for yolox model
label_path string ”” The label file with label names for detected objects written on it
precision string “fp16” The inference mode: “fp32”, “fp16”, “int8”
build_only bool false shutdown node after TensorRT engine file is built
calibration_algorithm string “MinMax” Calibration algorithm to be used for quantization when precision==int8. Valid value is one of: Entropy”,(“Legacy” | “Percentile”), “MinMax”]
dla_core_id int -1 If positive ID value is specified, the node assign inference task to the DLA core
quantize_first_layer bool false If true, set the operating precision for the first (input) layer to be fp16. This option is valid only when precision==int8
quantize_last_layer bool false If true, set the operating precision for the last (output) layer to be fp16. This option is valid only when precision==int8
profile_per_layer bool false If true, profiler function will be enabled. Since the profile function may affect execution speed, it is recommended to set this flag true only for development purpose.
clip_value double 0.0 If positive value is specified, the value of each layer output will be clipped between [0.0, clip_value]. This option is valid only when precision==int8 and used to manually specify the dynamic range instead of using any calibration
preprocess_on_gpu bool true If true, pre-processing is performed on GPU
calibration_image_list_path string ”” Path to a file which contains path to images. Those images will be used for int8 quantization.

Assumptions / Known limits

The label contained in detected 2D bounding boxes (i.e., out/objects) will be either one of the followings:

  • CAR
  • PEDESTRIAN (“PERSON” will also be categorized as “PEDESTRIAN”)
  • BUS
  • TRUCK
  • BICYCLE
  • MOTORCYCLE

If other labels (case insensitive) are contained in the file specified via the label_file parameter, those are labeled as UNKNOWN, while detected rectangles are drawn in the visualization result (out/image).

Onnx model

A sample model (named yolox-tiny.onnx) is downloaded by ansible script on env preparation stage, if not, please, follow Manual downloading of artifacts. To accelerate Non-maximum-suppression (NMS), which is one of the common post-process after object detection inference, EfficientNMS_TRT module is attached after the ordinal YOLOX (tiny) network. The EfficientNMS_TRT module contains fixed values for score_threshold and nms_threshold in it, hence these parameters are ignored when users specify ONNX models including this module.

This package accepts both EfficientNMS_TRT attached ONNXs and models published from the official YOLOX repository (we referred to them as “plain” models).

In addition to yolox-tiny.onnx, a custom model named yolox-sPlus-opt.onnx is either available. This model is based on YOLOX-s and tuned to perform more accurate detection with almost comparable execution speed with yolox-tiny. To get better results with this model, users are recommended to use some specific running arguments such as precision:=int8, calibration_algorithm:=Entropy, clip_value:=6.0. Users can refer launch/yolox_sPlus_opt.launch.xml to see how this model can be used.

All models are automatically converted to TensorRT format. These converted files will be saved in the same directory as specified ONNX files with .engine filename extension and reused from the next run. The conversion process may take a while (typically 10 to 20 minutes) and the inference process is blocked until complete the conversion, so it will take some time until detection results are published (even until appearing in the topic list) on the first run

Package acceptable model generation

To convert users’ own model that saved in PyTorch’s pth format into ONNX, users can exploit the converter offered by the official repository. For the convenience, only procedures are described below. Please refer the official document for more detail.

File truncated at 100 lines see the full file

CHANGELOG
No CHANGELOG found.

Launch files

  • launch/yolox.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_s_plus_opt.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_tiny.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-tiny]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: fp16]
      • calibration_algorithm [default: MinMax]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 0.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged tensorrt_yolox at Robotics Stack Exchange

No version for distro jazzy showing github. Known supported distros are highlighted in the buttons above.
Package symbol

tensorrt_yolox package from autodrrt repo

autonomous_emergency_braking control_performance_analysis control_validator external_cmd_selector joy_controller lane_departure_checker mpc_lateral_controller obstacle_collision_checker operation_mode_transition_manager pid_longitudinal_controller predicted_path_checker pure_pursuit shift_decider trajectory_follower_base trajectory_follower_node vehicle_cmd_gate diagnostic_converter kinematic_evaluator localization_evaluator planning_evaluator ekf_localizer geo_pose_projector gyro_odometer ar_tag_based_localizer landmark_manager localization_error_monitor localization_gnss localization_util ndt_scan_matcher pose2twist pose_initializer pose_instability_detector stop_filter tree_structured_parzen_estimator twist2accel yabloc_common yabloc_image_processing yabloc_monitor yabloc_particle_filter yabloc_pose_initializer map_height_fitter map_loader map_projection_loader map_tf_generator lanelet2_map_preprocessor ros2_bevdet bytetrack cluster_merger compare_map_segmentation crosswalk_traffic_light_estimator detected_object_feature_remover detected_object_validation detection_by_tracker elevation_map_loader euclidean_cluster front_vehicle_velocity_estimator ground_segmentation heatmap_visualizer image_projection_based_fusion lidar_apollo_instance_segmentation lidar_apollo_segmentation_tvm lidar_apollo_segmentation_tvm_nodes lidar_centerpoint lidar_centerpoint_tvm map_based_prediction multi_object_tracker object_merger object_range_splitter object_velocity_splitter occupancy_grid_map_outlier_filter probabilistic_occupancy_grid_map radar_crossing_objects_noise_filter radar_fusion_to_detected_object radar_object_clustering radar_object_tracker radar_tracks_msgs_converter shape_estimation simple_object_merger tensorrt_classifier tensorrt_yolo tensorrt_yolox tracking_object_merger traffic_light_arbiter traffic_light_classifier traffic_light_fine_detector traffic_light_map_based_detector traffic_light_multi_camera_fusion traffic_light_occlusion_predictor traffic_light_ssd_fine_detector traffic_light_visualization behavior_path_avoidance_by_lane_change_module behavior_path_avoidance_module behavior_path_external_request_lane_change_module behavior_path_goal_planner_module behavior_path_lane_change_module behavior_path_planner behavior_path_planner_common behavior_path_side_shift_module behavior_path_start_planner_module behavior_velocity_blind_spot_module behavior_velocity_crosswalk_module behavior_velocity_detection_area_module behavior_velocity_intersection_module behavior_velocity_no_drivable_lane_module behavior_velocity_no_stopping_area_module behavior_velocity_occlusion_spot_module behavior_velocity_out_of_lane_module behavior_velocity_planner behavior_velocity_planner_common behavior_velocity_run_out_module behavior_velocity_speed_bump_module behavior_velocity_stop_line_module behavior_velocity_template_module behavior_velocity_traffic_light_module behavior_velocity_virtual_traffic_light_module behavior_velocity_walkway_module costmap_generator external_velocity_limit_selector freespace_planner freespace_planning_algorithms mission_planner motion_velocity_smoother objects_of_interest_marker_interface obstacle_avoidance_planner obstacle_cruise_planner obstacle_stop_planner obstacle_velocity_limiter path_smoother planning_debug_tools planning_test_utils planning_topic_converter planning_validator route_handler rtc_interface rtc_replayer bezier_sampler frenet_planner path_sampler sampler_common scenario_selector static_centerline_optimizer surround_obstacle_checker gnss_poser image_diagnostics image_transport_decompressor imu_corrector livox_tag_filter pointcloud_preprocessor radar_scan_to_pointcloud2 radar_static_pointcloud_filter radar_threshold_filter radar_tracks_noise_filter tier4_pcl_extensions vehicle_velocity_converter autoware_auto_msgs_adapter bluetooth_monitor component_state_monitor default_ad_api ad_api_adaptors ad_api_visualizers automatic_pose_initializer diagnostic_graph_aggregator dummy_diag_publisher dummy_infrastructure duplicated_node_checker emergency_handler mrm_comfortable_stop_operator mrm_emergency_stop_operator system_error_monitor system_monitor topic_state_monitor velodyne_monitor accel_brake_map_calibrator external_cmd_converter raw_vehicle_cmd_converter steer_offset_estimator vehicle_info_util launch loop_sim ssh_machine gpudirect-dds autoware_ad_api_specs autoware_adapi_v1_msgs autoware_adapi_version_msgs autoware_auto_common autoware_auto_geometry autoware_auto_control_msgs autoware_auto_geometry_msgs autoware_auto_mapping_msgs autoware_auto_msgs autoware_auto_perception_msgs autoware_auto_planning_msgs autoware_auto_system_msgs autoware_auto_vehicle_msgs autoware_auto_perception_rviz_plugin autoware_auto_tf2 autoware_cmake autoware_lint_common autoware_utils lanelet2_extension autoware_common_msgs autoware_control_msgs autoware_localization_msgs autoware_map_msgs autoware_perception_msgs autoware_planning_msgs autoware_sensing_msgs autoware_system_msgs autoware_vehicle_msgs autoware_point_types autoware_testing bag_time_manager_rviz_plugin component_interface_specs component_interface_tools component_interface_utils cuda_utils fake_test_node geography_utils global_parameter_loader glog_component goal_distance_calculator grid_map_utils interpolation kalman_filter motion_utils object_recognition_utils osqp_interface path_distance_calculator perception_utils polar_grid qp_interface rtc_manager_rviz_plugin signal_processing tensorrt_common tier4_adapi_rviz_plugin tier4_api_utils tier4_automatic_goal_rviz_plugin tier4_autoware_utils tier4_calibration_rviz_plugin tier4_camera_view_rviz_plugin tier4_control_rviz_plugin tier4_datetime_rviz_plugin tier4_debug_rviz_plugin tier4_debug_tools tier4_localization_rviz_plugin tier4_perception_rviz_plugin tier4_planning_rviz_plugin tier4_screen_capture_rviz_plugin tier4_simulated_clock_rviz_plugin tier4_state_rviz_plugin tier4_system_rviz_plugin tier4_target_object_type_rviz_plugin tier4_traffic_light_rviz_plugin tier4_vehicle_rviz_plugin time_utils simulator_compatibility_test traffic_light_recognition_marker_publisher traffic_light_utils tvm_utility dma_customer_msg dma_transfer eagleye_coordinate eagleye_navigation eagleye_msgs eagleye_rt eagleye_can_velocity_converter eagleye_fix2kml eagleye_geo_pose_converter eagleye_geo_pose_fusion eagleye_gnss_converter eagleye_tf llh_converter morai_msgs mussp ndt_omp orocos_kdl python_orocos_kdl pointcloud_to_laserscan rtklib_bridge rtklib_msgs autoware_external_api_msgs autoware_iv_external_api_adaptor autoware_iv_internal_api_adaptor awapi_awiv_adapter tier4_api_msgs tier4_auto_msgs_converter tier4_control_msgs tier4_debug_msgs tier4_external_api_msgs tier4_hmi_msgs tier4_localization_msgs tier4_map_msgs tier4_perception_msgs tier4_planning_msgs tier4_rtc_msgs tier4_simulation_msgs tier4_system_msgs tier4_v2x_msgs tier4_vehicle_msgs tier4_autoware_api_launch tier4_control_launch tier4_localization_launch tier4_map_launch tier4_perception_launch tier4_planning_launch tier4_sensing_launch tier4_simulator_launch tier4_system_launch tier4_vehicle_launch fastrtps cyclonedds lanelet2 lanelet2_core lanelet2_examples lanelet2_io lanelet2_maps lanelet2_matching lanelet2_projection lanelet2_python lanelet2_routing lanelet2_traffic_rules lanelet2_validation sophus angles behaviortree_cpp_v3 bond bond_core bondcpp bondpy smclib test_bond cudnn_cmake_module diagnostic_aggregator diagnostic_common_diagnostics diagnostic_updater diagnostics self_test filters geodesy geographic_info geographic_msgs grid_map grid_map_cmake_helpers grid_map_core grid_map_costmap_2d grid_map_cv grid_map_demos grid_map_filters grid_map_loader grid_map_msgs grid_map_octomap grid_map_pcl grid_map_ros grid_map_rviz_plugin grid_map_sdf grid_map_visualization mrt_cmake_modules nav2_amcl nav2_behavior_tree nav2_behaviors nav2_bringup nav2_bt_navigator nav2_collision_monitor nav2_common nav2_controller nav2_core nav2_costmap_2d costmap_queue dwb_core dwb_critics dwb_msgs dwb_plugins nav2_dwb_controller nav_2d_msgs nav_2d_utils nav2_lifecycle_manager nav2_map_server nav2_msgs nav2_navfn_planner nav2_planner nav2_regulated_pure_pursuit_controller nav2_rotation_shim_controller nav2_rviz_plugins nav2_simple_commander nav2_smac_planner nav2_smoother nav2_system_tests nav2_theta_star_planner nav2_util nav2_velocity_smoother nav2_voxel_grid nav2_waypoint_follower navigation2 dynamic_edt_3d octomap octovis octomap_msgs osqp_vendor pacmod3_msgs pcl_msgs pcl_conversions pcl_ros perception_pcl point_cloud_msg_wrapper radar_msgs can_msgs rqt_tf_tree tensorrt_cmake_module topic_tools topic_tools_interfaces tvm_vendor cv_bridge image_geometry opencv_tests vision_opencv xacro rviz2 rviz_assimp_vendor rviz_common rviz_default_plugins rviz_ogre_vendor rviz_rendering rviz_rendering_tests rviz_visual_testing_framework dummy_perception_publisher fault_injection simple_planning_simulator image_view v4l2_camera aion_interface carla_vehicle_can cgi430_can_driver cgi610_driver ARS408_driver common_sensor_launch kunyi_sensor_kit_description kunyi_sensor_kit_launch classformsg node_v2x data_format_dump data_preprocess_launch lidar_centerpoint_collect lidar_saver message_sync time_cal camera_calibration direct_visual_lidar_calibration multi_lidar_calibration

ROS Distro
github

API Docs
Browse Code

Package Summary

Tags No category tags.
Version 0.0.1
License Apache License 2.0
Build type AMENT_CMAKE
Use RECOMMENDED

Repository Summary

Description
Checkout URI https://github.com/ieiauto/autodrrt.git
VCS Type git
VCS Version main
Last Updated 2025-05-30
Dev Status UNKNOWN
Released UNRELEASED
Tags No category tags.
Contributing Help Wanted (-)
Good First Issues (-)
Pull Requests to Review (-)

Package Description

tensorrt library implementation for yolox

Additional Links

No additional links.

Maintainers

  • Daisuke Nishimatsu
  • Dan Umeda
  • Manato Hirabayashi

Authors

  • Daisuke Nishimatsu
src/autodrrt.application/perception/autoware.perception/tensorrt_yolox/README.md

tensorrt_yolox

Purpose

This package detects target objects e.g., cars, trucks, bicycles, and pedestrians on a image based on YOLOX model.

Inner-workings / Algorithms

Cite

Zheng Ge, Songtao Liu, Feng Wang, Zeming Li, Jian Sun, “YOLOX: Exceeding YOLO Series in 2021”, arXiv preprint arXiv:2107.08430, 2021 [ref]

Inputs / Outputs

Input

Name Type Description
in/image sensor_msgs/Image The input image

Output

Name Type Description
out/objects tier4_perception_msgs/DetectedObjectsWithFeature The detected objects with 2D bounding boxes
out/image sensor_msgs/Image The image with 2D bounding boxes for visualization

Parameters

Core Parameters

Name Type Default Value Description
score_threshold float 0.3 If the objectness score is less than this value, the object is ignored in yolox layer.
nms_threshold float 0.7 The IoU threshold for NMS method

NOTE: These two parameters are only valid for “plain” model (described later).

Node Parameters

Name Type Default Value Description
model_path string ”” The onnx file name for yolox model
label_path string ”” The label file with label names for detected objects written on it
precision string “fp16” The inference mode: “fp32”, “fp16”, “int8”
build_only bool false shutdown node after TensorRT engine file is built
calibration_algorithm string “MinMax” Calibration algorithm to be used for quantization when precision==int8. Valid value is one of: Entropy”,(“Legacy” | “Percentile”), “MinMax”]
dla_core_id int -1 If positive ID value is specified, the node assign inference task to the DLA core
quantize_first_layer bool false If true, set the operating precision for the first (input) layer to be fp16. This option is valid only when precision==int8
quantize_last_layer bool false If true, set the operating precision for the last (output) layer to be fp16. This option is valid only when precision==int8
profile_per_layer bool false If true, profiler function will be enabled. Since the profile function may affect execution speed, it is recommended to set this flag true only for development purpose.
clip_value double 0.0 If positive value is specified, the value of each layer output will be clipped between [0.0, clip_value]. This option is valid only when precision==int8 and used to manually specify the dynamic range instead of using any calibration
preprocess_on_gpu bool true If true, pre-processing is performed on GPU
calibration_image_list_path string ”” Path to a file which contains path to images. Those images will be used for int8 quantization.

Assumptions / Known limits

The label contained in detected 2D bounding boxes (i.e., out/objects) will be either one of the followings:

  • CAR
  • PEDESTRIAN (“PERSON” will also be categorized as “PEDESTRIAN”)
  • BUS
  • TRUCK
  • BICYCLE
  • MOTORCYCLE

If other labels (case insensitive) are contained in the file specified via the label_file parameter, those are labeled as UNKNOWN, while detected rectangles are drawn in the visualization result (out/image).

Onnx model

A sample model (named yolox-tiny.onnx) is downloaded by ansible script on env preparation stage, if not, please, follow Manual downloading of artifacts. To accelerate Non-maximum-suppression (NMS), which is one of the common post-process after object detection inference, EfficientNMS_TRT module is attached after the ordinal YOLOX (tiny) network. The EfficientNMS_TRT module contains fixed values for score_threshold and nms_threshold in it, hence these parameters are ignored when users specify ONNX models including this module.

This package accepts both EfficientNMS_TRT attached ONNXs and models published from the official YOLOX repository (we referred to them as “plain” models).

In addition to yolox-tiny.onnx, a custom model named yolox-sPlus-opt.onnx is either available. This model is based on YOLOX-s and tuned to perform more accurate detection with almost comparable execution speed with yolox-tiny. To get better results with this model, users are recommended to use some specific running arguments such as precision:=int8, calibration_algorithm:=Entropy, clip_value:=6.0. Users can refer launch/yolox_sPlus_opt.launch.xml to see how this model can be used.

All models are automatically converted to TensorRT format. These converted files will be saved in the same directory as specified ONNX files with .engine filename extension and reused from the next run. The conversion process may take a while (typically 10 to 20 minutes) and the inference process is blocked until complete the conversion, so it will take some time until detection results are published (even until appearing in the topic list) on the first run

Package acceptable model generation

To convert users’ own model that saved in PyTorch’s pth format into ONNX, users can exploit the converter offered by the official repository. For the convenience, only procedures are described below. Please refer the official document for more detail.

File truncated at 100 lines see the full file

CHANGELOG
No CHANGELOG found.

Launch files

  • launch/yolox.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_s_plus_opt.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_tiny.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-tiny]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: fp16]
      • calibration_algorithm [default: MinMax]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 0.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged tensorrt_yolox at Robotics Stack Exchange

No version for distro kilted showing github. Known supported distros are highlighted in the buttons above.
Package symbol

tensorrt_yolox package from autodrrt repo

autonomous_emergency_braking control_performance_analysis control_validator external_cmd_selector joy_controller lane_departure_checker mpc_lateral_controller obstacle_collision_checker operation_mode_transition_manager pid_longitudinal_controller predicted_path_checker pure_pursuit shift_decider trajectory_follower_base trajectory_follower_node vehicle_cmd_gate diagnostic_converter kinematic_evaluator localization_evaluator planning_evaluator ekf_localizer geo_pose_projector gyro_odometer ar_tag_based_localizer landmark_manager localization_error_monitor localization_gnss localization_util ndt_scan_matcher pose2twist pose_initializer pose_instability_detector stop_filter tree_structured_parzen_estimator twist2accel yabloc_common yabloc_image_processing yabloc_monitor yabloc_particle_filter yabloc_pose_initializer map_height_fitter map_loader map_projection_loader map_tf_generator lanelet2_map_preprocessor ros2_bevdet bytetrack cluster_merger compare_map_segmentation crosswalk_traffic_light_estimator detected_object_feature_remover detected_object_validation detection_by_tracker elevation_map_loader euclidean_cluster front_vehicle_velocity_estimator ground_segmentation heatmap_visualizer image_projection_based_fusion lidar_apollo_instance_segmentation lidar_apollo_segmentation_tvm lidar_apollo_segmentation_tvm_nodes lidar_centerpoint lidar_centerpoint_tvm map_based_prediction multi_object_tracker object_merger object_range_splitter object_velocity_splitter occupancy_grid_map_outlier_filter probabilistic_occupancy_grid_map radar_crossing_objects_noise_filter radar_fusion_to_detected_object radar_object_clustering radar_object_tracker radar_tracks_msgs_converter shape_estimation simple_object_merger tensorrt_classifier tensorrt_yolo tensorrt_yolox tracking_object_merger traffic_light_arbiter traffic_light_classifier traffic_light_fine_detector traffic_light_map_based_detector traffic_light_multi_camera_fusion traffic_light_occlusion_predictor traffic_light_ssd_fine_detector traffic_light_visualization behavior_path_avoidance_by_lane_change_module behavior_path_avoidance_module behavior_path_external_request_lane_change_module behavior_path_goal_planner_module behavior_path_lane_change_module behavior_path_planner behavior_path_planner_common behavior_path_side_shift_module behavior_path_start_planner_module behavior_velocity_blind_spot_module behavior_velocity_crosswalk_module behavior_velocity_detection_area_module behavior_velocity_intersection_module behavior_velocity_no_drivable_lane_module behavior_velocity_no_stopping_area_module behavior_velocity_occlusion_spot_module behavior_velocity_out_of_lane_module behavior_velocity_planner behavior_velocity_planner_common behavior_velocity_run_out_module behavior_velocity_speed_bump_module behavior_velocity_stop_line_module behavior_velocity_template_module behavior_velocity_traffic_light_module behavior_velocity_virtual_traffic_light_module behavior_velocity_walkway_module costmap_generator external_velocity_limit_selector freespace_planner freespace_planning_algorithms mission_planner motion_velocity_smoother objects_of_interest_marker_interface obstacle_avoidance_planner obstacle_cruise_planner obstacle_stop_planner obstacle_velocity_limiter path_smoother planning_debug_tools planning_test_utils planning_topic_converter planning_validator route_handler rtc_interface rtc_replayer bezier_sampler frenet_planner path_sampler sampler_common scenario_selector static_centerline_optimizer surround_obstacle_checker gnss_poser image_diagnostics image_transport_decompressor imu_corrector livox_tag_filter pointcloud_preprocessor radar_scan_to_pointcloud2 radar_static_pointcloud_filter radar_threshold_filter radar_tracks_noise_filter tier4_pcl_extensions vehicle_velocity_converter autoware_auto_msgs_adapter bluetooth_monitor component_state_monitor default_ad_api ad_api_adaptors ad_api_visualizers automatic_pose_initializer diagnostic_graph_aggregator dummy_diag_publisher dummy_infrastructure duplicated_node_checker emergency_handler mrm_comfortable_stop_operator mrm_emergency_stop_operator system_error_monitor system_monitor topic_state_monitor velodyne_monitor accel_brake_map_calibrator external_cmd_converter raw_vehicle_cmd_converter steer_offset_estimator vehicle_info_util launch loop_sim ssh_machine gpudirect-dds autoware_ad_api_specs autoware_adapi_v1_msgs autoware_adapi_version_msgs autoware_auto_common autoware_auto_geometry autoware_auto_control_msgs autoware_auto_geometry_msgs autoware_auto_mapping_msgs autoware_auto_msgs autoware_auto_perception_msgs autoware_auto_planning_msgs autoware_auto_system_msgs autoware_auto_vehicle_msgs autoware_auto_perception_rviz_plugin autoware_auto_tf2 autoware_cmake autoware_lint_common autoware_utils lanelet2_extension autoware_common_msgs autoware_control_msgs autoware_localization_msgs autoware_map_msgs autoware_perception_msgs autoware_planning_msgs autoware_sensing_msgs autoware_system_msgs autoware_vehicle_msgs autoware_point_types autoware_testing bag_time_manager_rviz_plugin component_interface_specs component_interface_tools component_interface_utils cuda_utils fake_test_node geography_utils global_parameter_loader glog_component goal_distance_calculator grid_map_utils interpolation kalman_filter motion_utils object_recognition_utils osqp_interface path_distance_calculator perception_utils polar_grid qp_interface rtc_manager_rviz_plugin signal_processing tensorrt_common tier4_adapi_rviz_plugin tier4_api_utils tier4_automatic_goal_rviz_plugin tier4_autoware_utils tier4_calibration_rviz_plugin tier4_camera_view_rviz_plugin tier4_control_rviz_plugin tier4_datetime_rviz_plugin tier4_debug_rviz_plugin tier4_debug_tools tier4_localization_rviz_plugin tier4_perception_rviz_plugin tier4_planning_rviz_plugin tier4_screen_capture_rviz_plugin tier4_simulated_clock_rviz_plugin tier4_state_rviz_plugin tier4_system_rviz_plugin tier4_target_object_type_rviz_plugin tier4_traffic_light_rviz_plugin tier4_vehicle_rviz_plugin time_utils simulator_compatibility_test traffic_light_recognition_marker_publisher traffic_light_utils tvm_utility dma_customer_msg dma_transfer eagleye_coordinate eagleye_navigation eagleye_msgs eagleye_rt eagleye_can_velocity_converter eagleye_fix2kml eagleye_geo_pose_converter eagleye_geo_pose_fusion eagleye_gnss_converter eagleye_tf llh_converter morai_msgs mussp ndt_omp orocos_kdl python_orocos_kdl pointcloud_to_laserscan rtklib_bridge rtklib_msgs autoware_external_api_msgs autoware_iv_external_api_adaptor autoware_iv_internal_api_adaptor awapi_awiv_adapter tier4_api_msgs tier4_auto_msgs_converter tier4_control_msgs tier4_debug_msgs tier4_external_api_msgs tier4_hmi_msgs tier4_localization_msgs tier4_map_msgs tier4_perception_msgs tier4_planning_msgs tier4_rtc_msgs tier4_simulation_msgs tier4_system_msgs tier4_v2x_msgs tier4_vehicle_msgs tier4_autoware_api_launch tier4_control_launch tier4_localization_launch tier4_map_launch tier4_perception_launch tier4_planning_launch tier4_sensing_launch tier4_simulator_launch tier4_system_launch tier4_vehicle_launch fastrtps cyclonedds lanelet2 lanelet2_core lanelet2_examples lanelet2_io lanelet2_maps lanelet2_matching lanelet2_projection lanelet2_python lanelet2_routing lanelet2_traffic_rules lanelet2_validation sophus angles behaviortree_cpp_v3 bond bond_core bondcpp bondpy smclib test_bond cudnn_cmake_module diagnostic_aggregator diagnostic_common_diagnostics diagnostic_updater diagnostics self_test filters geodesy geographic_info geographic_msgs grid_map grid_map_cmake_helpers grid_map_core grid_map_costmap_2d grid_map_cv grid_map_demos grid_map_filters grid_map_loader grid_map_msgs grid_map_octomap grid_map_pcl grid_map_ros grid_map_rviz_plugin grid_map_sdf grid_map_visualization mrt_cmake_modules nav2_amcl nav2_behavior_tree nav2_behaviors nav2_bringup nav2_bt_navigator nav2_collision_monitor nav2_common nav2_controller nav2_core nav2_costmap_2d costmap_queue dwb_core dwb_critics dwb_msgs dwb_plugins nav2_dwb_controller nav_2d_msgs nav_2d_utils nav2_lifecycle_manager nav2_map_server nav2_msgs nav2_navfn_planner nav2_planner nav2_regulated_pure_pursuit_controller nav2_rotation_shim_controller nav2_rviz_plugins nav2_simple_commander nav2_smac_planner nav2_smoother nav2_system_tests nav2_theta_star_planner nav2_util nav2_velocity_smoother nav2_voxel_grid nav2_waypoint_follower navigation2 dynamic_edt_3d octomap octovis octomap_msgs osqp_vendor pacmod3_msgs pcl_msgs pcl_conversions pcl_ros perception_pcl point_cloud_msg_wrapper radar_msgs can_msgs rqt_tf_tree tensorrt_cmake_module topic_tools topic_tools_interfaces tvm_vendor cv_bridge image_geometry opencv_tests vision_opencv xacro rviz2 rviz_assimp_vendor rviz_common rviz_default_plugins rviz_ogre_vendor rviz_rendering rviz_rendering_tests rviz_visual_testing_framework dummy_perception_publisher fault_injection simple_planning_simulator image_view v4l2_camera aion_interface carla_vehicle_can cgi430_can_driver cgi610_driver ARS408_driver common_sensor_launch kunyi_sensor_kit_description kunyi_sensor_kit_launch classformsg node_v2x data_format_dump data_preprocess_launch lidar_centerpoint_collect lidar_saver message_sync time_cal camera_calibration direct_visual_lidar_calibration multi_lidar_calibration

ROS Distro
github

API Docs
Browse Code

Package Summary

Tags No category tags.
Version 0.0.1
License Apache License 2.0
Build type AMENT_CMAKE
Use RECOMMENDED

Repository Summary

Description
Checkout URI https://github.com/ieiauto/autodrrt.git
VCS Type git
VCS Version main
Last Updated 2025-05-30
Dev Status UNKNOWN
Released UNRELEASED
Tags No category tags.
Contributing Help Wanted (-)
Good First Issues (-)
Pull Requests to Review (-)

Package Description

tensorrt library implementation for yolox

Additional Links

No additional links.

Maintainers

  • Daisuke Nishimatsu
  • Dan Umeda
  • Manato Hirabayashi

Authors

  • Daisuke Nishimatsu
src/autodrrt.application/perception/autoware.perception/tensorrt_yolox/README.md

tensorrt_yolox

Purpose

This package detects target objects e.g., cars, trucks, bicycles, and pedestrians on a image based on YOLOX model.

Inner-workings / Algorithms

Cite

Zheng Ge, Songtao Liu, Feng Wang, Zeming Li, Jian Sun, “YOLOX: Exceeding YOLO Series in 2021”, arXiv preprint arXiv:2107.08430, 2021 [ref]

Inputs / Outputs

Input

Name Type Description
in/image sensor_msgs/Image The input image

Output

Name Type Description
out/objects tier4_perception_msgs/DetectedObjectsWithFeature The detected objects with 2D bounding boxes
out/image sensor_msgs/Image The image with 2D bounding boxes for visualization

Parameters

Core Parameters

Name Type Default Value Description
score_threshold float 0.3 If the objectness score is less than this value, the object is ignored in yolox layer.
nms_threshold float 0.7 The IoU threshold for NMS method

NOTE: These two parameters are only valid for “plain” model (described later).

Node Parameters

Name Type Default Value Description
model_path string ”” The onnx file name for yolox model
label_path string ”” The label file with label names for detected objects written on it
precision string “fp16” The inference mode: “fp32”, “fp16”, “int8”
build_only bool false shutdown node after TensorRT engine file is built
calibration_algorithm string “MinMax” Calibration algorithm to be used for quantization when precision==int8. Valid value is one of: Entropy”,(“Legacy” | “Percentile”), “MinMax”]
dla_core_id int -1 If positive ID value is specified, the node assign inference task to the DLA core
quantize_first_layer bool false If true, set the operating precision for the first (input) layer to be fp16. This option is valid only when precision==int8
quantize_last_layer bool false If true, set the operating precision for the last (output) layer to be fp16. This option is valid only when precision==int8
profile_per_layer bool false If true, profiler function will be enabled. Since the profile function may affect execution speed, it is recommended to set this flag true only for development purpose.
clip_value double 0.0 If positive value is specified, the value of each layer output will be clipped between [0.0, clip_value]. This option is valid only when precision==int8 and used to manually specify the dynamic range instead of using any calibration
preprocess_on_gpu bool true If true, pre-processing is performed on GPU
calibration_image_list_path string ”” Path to a file which contains path to images. Those images will be used for int8 quantization.

Assumptions / Known limits

The label contained in detected 2D bounding boxes (i.e., out/objects) will be either one of the followings:

  • CAR
  • PEDESTRIAN (“PERSON” will also be categorized as “PEDESTRIAN”)
  • BUS
  • TRUCK
  • BICYCLE
  • MOTORCYCLE

If other labels (case insensitive) are contained in the file specified via the label_file parameter, those are labeled as UNKNOWN, while detected rectangles are drawn in the visualization result (out/image).

Onnx model

A sample model (named yolox-tiny.onnx) is downloaded by ansible script on env preparation stage, if not, please, follow Manual downloading of artifacts. To accelerate Non-maximum-suppression (NMS), which is one of the common post-process after object detection inference, EfficientNMS_TRT module is attached after the ordinal YOLOX (tiny) network. The EfficientNMS_TRT module contains fixed values for score_threshold and nms_threshold in it, hence these parameters are ignored when users specify ONNX models including this module.

This package accepts both EfficientNMS_TRT attached ONNXs and models published from the official YOLOX repository (we referred to them as “plain” models).

In addition to yolox-tiny.onnx, a custom model named yolox-sPlus-opt.onnx is either available. This model is based on YOLOX-s and tuned to perform more accurate detection with almost comparable execution speed with yolox-tiny. To get better results with this model, users are recommended to use some specific running arguments such as precision:=int8, calibration_algorithm:=Entropy, clip_value:=6.0. Users can refer launch/yolox_sPlus_opt.launch.xml to see how this model can be used.

All models are automatically converted to TensorRT format. These converted files will be saved in the same directory as specified ONNX files with .engine filename extension and reused from the next run. The conversion process may take a while (typically 10 to 20 minutes) and the inference process is blocked until complete the conversion, so it will take some time until detection results are published (even until appearing in the topic list) on the first run

Package acceptable model generation

To convert users’ own model that saved in PyTorch’s pth format into ONNX, users can exploit the converter offered by the official repository. For the convenience, only procedures are described below. Please refer the official document for more detail.

File truncated at 100 lines see the full file

CHANGELOG
No CHANGELOG found.

Launch files

  • launch/yolox.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_s_plus_opt.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_tiny.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-tiny]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: fp16]
      • calibration_algorithm [default: MinMax]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 0.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged tensorrt_yolox at Robotics Stack Exchange

No version for distro rolling showing github. Known supported distros are highlighted in the buttons above.
Package symbol

tensorrt_yolox package from autodrrt repo

autonomous_emergency_braking control_performance_analysis control_validator external_cmd_selector joy_controller lane_departure_checker mpc_lateral_controller obstacle_collision_checker operation_mode_transition_manager pid_longitudinal_controller predicted_path_checker pure_pursuit shift_decider trajectory_follower_base trajectory_follower_node vehicle_cmd_gate diagnostic_converter kinematic_evaluator localization_evaluator planning_evaluator ekf_localizer geo_pose_projector gyro_odometer ar_tag_based_localizer landmark_manager localization_error_monitor localization_gnss localization_util ndt_scan_matcher pose2twist pose_initializer pose_instability_detector stop_filter tree_structured_parzen_estimator twist2accel yabloc_common yabloc_image_processing yabloc_monitor yabloc_particle_filter yabloc_pose_initializer map_height_fitter map_loader map_projection_loader map_tf_generator lanelet2_map_preprocessor ros2_bevdet bytetrack cluster_merger compare_map_segmentation crosswalk_traffic_light_estimator detected_object_feature_remover detected_object_validation detection_by_tracker elevation_map_loader euclidean_cluster front_vehicle_velocity_estimator ground_segmentation heatmap_visualizer image_projection_based_fusion lidar_apollo_instance_segmentation lidar_apollo_segmentation_tvm lidar_apollo_segmentation_tvm_nodes lidar_centerpoint lidar_centerpoint_tvm map_based_prediction multi_object_tracker object_merger object_range_splitter object_velocity_splitter occupancy_grid_map_outlier_filter probabilistic_occupancy_grid_map radar_crossing_objects_noise_filter radar_fusion_to_detected_object radar_object_clustering radar_object_tracker radar_tracks_msgs_converter shape_estimation simple_object_merger tensorrt_classifier tensorrt_yolo tensorrt_yolox tracking_object_merger traffic_light_arbiter traffic_light_classifier traffic_light_fine_detector traffic_light_map_based_detector traffic_light_multi_camera_fusion traffic_light_occlusion_predictor traffic_light_ssd_fine_detector traffic_light_visualization behavior_path_avoidance_by_lane_change_module behavior_path_avoidance_module behavior_path_external_request_lane_change_module behavior_path_goal_planner_module behavior_path_lane_change_module behavior_path_planner behavior_path_planner_common behavior_path_side_shift_module behavior_path_start_planner_module behavior_velocity_blind_spot_module behavior_velocity_crosswalk_module behavior_velocity_detection_area_module behavior_velocity_intersection_module behavior_velocity_no_drivable_lane_module behavior_velocity_no_stopping_area_module behavior_velocity_occlusion_spot_module behavior_velocity_out_of_lane_module behavior_velocity_planner behavior_velocity_planner_common behavior_velocity_run_out_module behavior_velocity_speed_bump_module behavior_velocity_stop_line_module behavior_velocity_template_module behavior_velocity_traffic_light_module behavior_velocity_virtual_traffic_light_module behavior_velocity_walkway_module costmap_generator external_velocity_limit_selector freespace_planner freespace_planning_algorithms mission_planner motion_velocity_smoother objects_of_interest_marker_interface obstacle_avoidance_planner obstacle_cruise_planner obstacle_stop_planner obstacle_velocity_limiter path_smoother planning_debug_tools planning_test_utils planning_topic_converter planning_validator route_handler rtc_interface rtc_replayer bezier_sampler frenet_planner path_sampler sampler_common scenario_selector static_centerline_optimizer surround_obstacle_checker gnss_poser image_diagnostics image_transport_decompressor imu_corrector livox_tag_filter pointcloud_preprocessor radar_scan_to_pointcloud2 radar_static_pointcloud_filter radar_threshold_filter radar_tracks_noise_filter tier4_pcl_extensions vehicle_velocity_converter autoware_auto_msgs_adapter bluetooth_monitor component_state_monitor default_ad_api ad_api_adaptors ad_api_visualizers automatic_pose_initializer diagnostic_graph_aggregator dummy_diag_publisher dummy_infrastructure duplicated_node_checker emergency_handler mrm_comfortable_stop_operator mrm_emergency_stop_operator system_error_monitor system_monitor topic_state_monitor velodyne_monitor accel_brake_map_calibrator external_cmd_converter raw_vehicle_cmd_converter steer_offset_estimator vehicle_info_util launch loop_sim ssh_machine gpudirect-dds autoware_ad_api_specs autoware_adapi_v1_msgs autoware_adapi_version_msgs autoware_auto_common autoware_auto_geometry autoware_auto_control_msgs autoware_auto_geometry_msgs autoware_auto_mapping_msgs autoware_auto_msgs autoware_auto_perception_msgs autoware_auto_planning_msgs autoware_auto_system_msgs autoware_auto_vehicle_msgs autoware_auto_perception_rviz_plugin autoware_auto_tf2 autoware_cmake autoware_lint_common autoware_utils lanelet2_extension autoware_common_msgs autoware_control_msgs autoware_localization_msgs autoware_map_msgs autoware_perception_msgs autoware_planning_msgs autoware_sensing_msgs autoware_system_msgs autoware_vehicle_msgs autoware_point_types autoware_testing bag_time_manager_rviz_plugin component_interface_specs component_interface_tools component_interface_utils cuda_utils fake_test_node geography_utils global_parameter_loader glog_component goal_distance_calculator grid_map_utils interpolation kalman_filter motion_utils object_recognition_utils osqp_interface path_distance_calculator perception_utils polar_grid qp_interface rtc_manager_rviz_plugin signal_processing tensorrt_common tier4_adapi_rviz_plugin tier4_api_utils tier4_automatic_goal_rviz_plugin tier4_autoware_utils tier4_calibration_rviz_plugin tier4_camera_view_rviz_plugin tier4_control_rviz_plugin tier4_datetime_rviz_plugin tier4_debug_rviz_plugin tier4_debug_tools tier4_localization_rviz_plugin tier4_perception_rviz_plugin tier4_planning_rviz_plugin tier4_screen_capture_rviz_plugin tier4_simulated_clock_rviz_plugin tier4_state_rviz_plugin tier4_system_rviz_plugin tier4_target_object_type_rviz_plugin tier4_traffic_light_rviz_plugin tier4_vehicle_rviz_plugin time_utils simulator_compatibility_test traffic_light_recognition_marker_publisher traffic_light_utils tvm_utility dma_customer_msg dma_transfer eagleye_coordinate eagleye_navigation eagleye_msgs eagleye_rt eagleye_can_velocity_converter eagleye_fix2kml eagleye_geo_pose_converter eagleye_geo_pose_fusion eagleye_gnss_converter eagleye_tf llh_converter morai_msgs mussp ndt_omp orocos_kdl python_orocos_kdl pointcloud_to_laserscan rtklib_bridge rtklib_msgs autoware_external_api_msgs autoware_iv_external_api_adaptor autoware_iv_internal_api_adaptor awapi_awiv_adapter tier4_api_msgs tier4_auto_msgs_converter tier4_control_msgs tier4_debug_msgs tier4_external_api_msgs tier4_hmi_msgs tier4_localization_msgs tier4_map_msgs tier4_perception_msgs tier4_planning_msgs tier4_rtc_msgs tier4_simulation_msgs tier4_system_msgs tier4_v2x_msgs tier4_vehicle_msgs tier4_autoware_api_launch tier4_control_launch tier4_localization_launch tier4_map_launch tier4_perception_launch tier4_planning_launch tier4_sensing_launch tier4_simulator_launch tier4_system_launch tier4_vehicle_launch fastrtps cyclonedds lanelet2 lanelet2_core lanelet2_examples lanelet2_io lanelet2_maps lanelet2_matching lanelet2_projection lanelet2_python lanelet2_routing lanelet2_traffic_rules lanelet2_validation sophus angles behaviortree_cpp_v3 bond bond_core bondcpp bondpy smclib test_bond cudnn_cmake_module diagnostic_aggregator diagnostic_common_diagnostics diagnostic_updater diagnostics self_test filters geodesy geographic_info geographic_msgs grid_map grid_map_cmake_helpers grid_map_core grid_map_costmap_2d grid_map_cv grid_map_demos grid_map_filters grid_map_loader grid_map_msgs grid_map_octomap grid_map_pcl grid_map_ros grid_map_rviz_plugin grid_map_sdf grid_map_visualization mrt_cmake_modules nav2_amcl nav2_behavior_tree nav2_behaviors nav2_bringup nav2_bt_navigator nav2_collision_monitor nav2_common nav2_controller nav2_core nav2_costmap_2d costmap_queue dwb_core dwb_critics dwb_msgs dwb_plugins nav2_dwb_controller nav_2d_msgs nav_2d_utils nav2_lifecycle_manager nav2_map_server nav2_msgs nav2_navfn_planner nav2_planner nav2_regulated_pure_pursuit_controller nav2_rotation_shim_controller nav2_rviz_plugins nav2_simple_commander nav2_smac_planner nav2_smoother nav2_system_tests nav2_theta_star_planner nav2_util nav2_velocity_smoother nav2_voxel_grid nav2_waypoint_follower navigation2 dynamic_edt_3d octomap octovis octomap_msgs osqp_vendor pacmod3_msgs pcl_msgs pcl_conversions pcl_ros perception_pcl point_cloud_msg_wrapper radar_msgs can_msgs rqt_tf_tree tensorrt_cmake_module topic_tools topic_tools_interfaces tvm_vendor cv_bridge image_geometry opencv_tests vision_opencv xacro rviz2 rviz_assimp_vendor rviz_common rviz_default_plugins rviz_ogre_vendor rviz_rendering rviz_rendering_tests rviz_visual_testing_framework dummy_perception_publisher fault_injection simple_planning_simulator image_view v4l2_camera aion_interface carla_vehicle_can cgi430_can_driver cgi610_driver ARS408_driver common_sensor_launch kunyi_sensor_kit_description kunyi_sensor_kit_launch classformsg node_v2x data_format_dump data_preprocess_launch lidar_centerpoint_collect lidar_saver message_sync time_cal camera_calibration direct_visual_lidar_calibration multi_lidar_calibration

ROS Distro
github

API Docs
Browse Code

Package Summary

Tags No category tags.
Version 0.0.1
License Apache License 2.0
Build type AMENT_CMAKE
Use RECOMMENDED

Repository Summary

Description
Checkout URI https://github.com/ieiauto/autodrrt.git
VCS Type git
VCS Version main
Last Updated 2025-05-30
Dev Status UNKNOWN
Released UNRELEASED
Tags No category tags.
Contributing Help Wanted (-)
Good First Issues (-)
Pull Requests to Review (-)

Package Description

tensorrt library implementation for yolox

Additional Links

No additional links.

Maintainers

  • Daisuke Nishimatsu
  • Dan Umeda
  • Manato Hirabayashi

Authors

  • Daisuke Nishimatsu
src/autodrrt.application/perception/autoware.perception/tensorrt_yolox/README.md

tensorrt_yolox

Purpose

This package detects target objects e.g., cars, trucks, bicycles, and pedestrians on a image based on YOLOX model.

Inner-workings / Algorithms

Cite

Zheng Ge, Songtao Liu, Feng Wang, Zeming Li, Jian Sun, “YOLOX: Exceeding YOLO Series in 2021”, arXiv preprint arXiv:2107.08430, 2021 [ref]

Inputs / Outputs

Input

Name Type Description
in/image sensor_msgs/Image The input image

Output

Name Type Description
out/objects tier4_perception_msgs/DetectedObjectsWithFeature The detected objects with 2D bounding boxes
out/image sensor_msgs/Image The image with 2D bounding boxes for visualization

Parameters

Core Parameters

Name Type Default Value Description
score_threshold float 0.3 If the objectness score is less than this value, the object is ignored in yolox layer.
nms_threshold float 0.7 The IoU threshold for NMS method

NOTE: These two parameters are only valid for “plain” model (described later).

Node Parameters

Name Type Default Value Description
model_path string ”” The onnx file name for yolox model
label_path string ”” The label file with label names for detected objects written on it
precision string “fp16” The inference mode: “fp32”, “fp16”, “int8”
build_only bool false shutdown node after TensorRT engine file is built
calibration_algorithm string “MinMax” Calibration algorithm to be used for quantization when precision==int8. Valid value is one of: Entropy”,(“Legacy” | “Percentile”), “MinMax”]
dla_core_id int -1 If positive ID value is specified, the node assign inference task to the DLA core
quantize_first_layer bool false If true, set the operating precision for the first (input) layer to be fp16. This option is valid only when precision==int8
quantize_last_layer bool false If true, set the operating precision for the last (output) layer to be fp16. This option is valid only when precision==int8
profile_per_layer bool false If true, profiler function will be enabled. Since the profile function may affect execution speed, it is recommended to set this flag true only for development purpose.
clip_value double 0.0 If positive value is specified, the value of each layer output will be clipped between [0.0, clip_value]. This option is valid only when precision==int8 and used to manually specify the dynamic range instead of using any calibration
preprocess_on_gpu bool true If true, pre-processing is performed on GPU
calibration_image_list_path string ”” Path to a file which contains path to images. Those images will be used for int8 quantization.

Assumptions / Known limits

The label contained in detected 2D bounding boxes (i.e., out/objects) will be either one of the followings:

  • CAR
  • PEDESTRIAN (“PERSON” will also be categorized as “PEDESTRIAN”)
  • BUS
  • TRUCK
  • BICYCLE
  • MOTORCYCLE

If other labels (case insensitive) are contained in the file specified via the label_file parameter, those are labeled as UNKNOWN, while detected rectangles are drawn in the visualization result (out/image).

Onnx model

A sample model (named yolox-tiny.onnx) is downloaded by ansible script on env preparation stage, if not, please, follow Manual downloading of artifacts. To accelerate Non-maximum-suppression (NMS), which is one of the common post-process after object detection inference, EfficientNMS_TRT module is attached after the ordinal YOLOX (tiny) network. The EfficientNMS_TRT module contains fixed values for score_threshold and nms_threshold in it, hence these parameters are ignored when users specify ONNX models including this module.

This package accepts both EfficientNMS_TRT attached ONNXs and models published from the official YOLOX repository (we referred to them as “plain” models).

In addition to yolox-tiny.onnx, a custom model named yolox-sPlus-opt.onnx is either available. This model is based on YOLOX-s and tuned to perform more accurate detection with almost comparable execution speed with yolox-tiny. To get better results with this model, users are recommended to use some specific running arguments such as precision:=int8, calibration_algorithm:=Entropy, clip_value:=6.0. Users can refer launch/yolox_sPlus_opt.launch.xml to see how this model can be used.

All models are automatically converted to TensorRT format. These converted files will be saved in the same directory as specified ONNX files with .engine filename extension and reused from the next run. The conversion process may take a while (typically 10 to 20 minutes) and the inference process is blocked until complete the conversion, so it will take some time until detection results are published (even until appearing in the topic list) on the first run

Package acceptable model generation

To convert users’ own model that saved in PyTorch’s pth format into ONNX, users can exploit the converter offered by the official repository. For the convenience, only procedures are described below. Please refer the official document for more detail.

File truncated at 100 lines see the full file

CHANGELOG
No CHANGELOG found.

Launch files

  • launch/yolox.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_s_plus_opt.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_tiny.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-tiny]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: fp16]
      • calibration_algorithm [default: MinMax]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 0.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged tensorrt_yolox at Robotics Stack Exchange

Package symbol

tensorrt_yolox package from autodrrt repo

autonomous_emergency_braking control_performance_analysis control_validator external_cmd_selector joy_controller lane_departure_checker mpc_lateral_controller obstacle_collision_checker operation_mode_transition_manager pid_longitudinal_controller predicted_path_checker pure_pursuit shift_decider trajectory_follower_base trajectory_follower_node vehicle_cmd_gate diagnostic_converter kinematic_evaluator localization_evaluator planning_evaluator ekf_localizer geo_pose_projector gyro_odometer ar_tag_based_localizer landmark_manager localization_error_monitor localization_gnss localization_util ndt_scan_matcher pose2twist pose_initializer pose_instability_detector stop_filter tree_structured_parzen_estimator twist2accel yabloc_common yabloc_image_processing yabloc_monitor yabloc_particle_filter yabloc_pose_initializer map_height_fitter map_loader map_projection_loader map_tf_generator lanelet2_map_preprocessor ros2_bevdet bytetrack cluster_merger compare_map_segmentation crosswalk_traffic_light_estimator detected_object_feature_remover detected_object_validation detection_by_tracker elevation_map_loader euclidean_cluster front_vehicle_velocity_estimator ground_segmentation heatmap_visualizer image_projection_based_fusion lidar_apollo_instance_segmentation lidar_apollo_segmentation_tvm lidar_apollo_segmentation_tvm_nodes lidar_centerpoint lidar_centerpoint_tvm map_based_prediction multi_object_tracker object_merger object_range_splitter object_velocity_splitter occupancy_grid_map_outlier_filter probabilistic_occupancy_grid_map radar_crossing_objects_noise_filter radar_fusion_to_detected_object radar_object_clustering radar_object_tracker radar_tracks_msgs_converter shape_estimation simple_object_merger tensorrt_classifier tensorrt_yolo tensorrt_yolox tracking_object_merger traffic_light_arbiter traffic_light_classifier traffic_light_fine_detector traffic_light_map_based_detector traffic_light_multi_camera_fusion traffic_light_occlusion_predictor traffic_light_ssd_fine_detector traffic_light_visualization behavior_path_avoidance_by_lane_change_module behavior_path_avoidance_module behavior_path_external_request_lane_change_module behavior_path_goal_planner_module behavior_path_lane_change_module behavior_path_planner behavior_path_planner_common behavior_path_side_shift_module behavior_path_start_planner_module behavior_velocity_blind_spot_module behavior_velocity_crosswalk_module behavior_velocity_detection_area_module behavior_velocity_intersection_module behavior_velocity_no_drivable_lane_module behavior_velocity_no_stopping_area_module behavior_velocity_occlusion_spot_module behavior_velocity_out_of_lane_module behavior_velocity_planner behavior_velocity_planner_common behavior_velocity_run_out_module behavior_velocity_speed_bump_module behavior_velocity_stop_line_module behavior_velocity_template_module behavior_velocity_traffic_light_module behavior_velocity_virtual_traffic_light_module behavior_velocity_walkway_module costmap_generator external_velocity_limit_selector freespace_planner freespace_planning_algorithms mission_planner motion_velocity_smoother objects_of_interest_marker_interface obstacle_avoidance_planner obstacle_cruise_planner obstacle_stop_planner obstacle_velocity_limiter path_smoother planning_debug_tools planning_test_utils planning_topic_converter planning_validator route_handler rtc_interface rtc_replayer bezier_sampler frenet_planner path_sampler sampler_common scenario_selector static_centerline_optimizer surround_obstacle_checker gnss_poser image_diagnostics image_transport_decompressor imu_corrector livox_tag_filter pointcloud_preprocessor radar_scan_to_pointcloud2 radar_static_pointcloud_filter radar_threshold_filter radar_tracks_noise_filter tier4_pcl_extensions vehicle_velocity_converter autoware_auto_msgs_adapter bluetooth_monitor component_state_monitor default_ad_api ad_api_adaptors ad_api_visualizers automatic_pose_initializer diagnostic_graph_aggregator dummy_diag_publisher dummy_infrastructure duplicated_node_checker emergency_handler mrm_comfortable_stop_operator mrm_emergency_stop_operator system_error_monitor system_monitor topic_state_monitor velodyne_monitor accel_brake_map_calibrator external_cmd_converter raw_vehicle_cmd_converter steer_offset_estimator vehicle_info_util launch loop_sim ssh_machine gpudirect-dds autoware_ad_api_specs autoware_adapi_v1_msgs autoware_adapi_version_msgs autoware_auto_common autoware_auto_geometry autoware_auto_control_msgs autoware_auto_geometry_msgs autoware_auto_mapping_msgs autoware_auto_msgs autoware_auto_perception_msgs autoware_auto_planning_msgs autoware_auto_system_msgs autoware_auto_vehicle_msgs autoware_auto_perception_rviz_plugin autoware_auto_tf2 autoware_cmake autoware_lint_common autoware_utils lanelet2_extension autoware_common_msgs autoware_control_msgs autoware_localization_msgs autoware_map_msgs autoware_perception_msgs autoware_planning_msgs autoware_sensing_msgs autoware_system_msgs autoware_vehicle_msgs autoware_point_types autoware_testing bag_time_manager_rviz_plugin component_interface_specs component_interface_tools component_interface_utils cuda_utils fake_test_node geography_utils global_parameter_loader glog_component goal_distance_calculator grid_map_utils interpolation kalman_filter motion_utils object_recognition_utils osqp_interface path_distance_calculator perception_utils polar_grid qp_interface rtc_manager_rviz_plugin signal_processing tensorrt_common tier4_adapi_rviz_plugin tier4_api_utils tier4_automatic_goal_rviz_plugin tier4_autoware_utils tier4_calibration_rviz_plugin tier4_camera_view_rviz_plugin tier4_control_rviz_plugin tier4_datetime_rviz_plugin tier4_debug_rviz_plugin tier4_debug_tools tier4_localization_rviz_plugin tier4_perception_rviz_plugin tier4_planning_rviz_plugin tier4_screen_capture_rviz_plugin tier4_simulated_clock_rviz_plugin tier4_state_rviz_plugin tier4_system_rviz_plugin tier4_target_object_type_rviz_plugin tier4_traffic_light_rviz_plugin tier4_vehicle_rviz_plugin time_utils simulator_compatibility_test traffic_light_recognition_marker_publisher traffic_light_utils tvm_utility dma_customer_msg dma_transfer eagleye_coordinate eagleye_navigation eagleye_msgs eagleye_rt eagleye_can_velocity_converter eagleye_fix2kml eagleye_geo_pose_converter eagleye_geo_pose_fusion eagleye_gnss_converter eagleye_tf llh_converter morai_msgs mussp ndt_omp orocos_kdl python_orocos_kdl pointcloud_to_laserscan rtklib_bridge rtklib_msgs autoware_external_api_msgs autoware_iv_external_api_adaptor autoware_iv_internal_api_adaptor awapi_awiv_adapter tier4_api_msgs tier4_auto_msgs_converter tier4_control_msgs tier4_debug_msgs tier4_external_api_msgs tier4_hmi_msgs tier4_localization_msgs tier4_map_msgs tier4_perception_msgs tier4_planning_msgs tier4_rtc_msgs tier4_simulation_msgs tier4_system_msgs tier4_v2x_msgs tier4_vehicle_msgs tier4_autoware_api_launch tier4_control_launch tier4_localization_launch tier4_map_launch tier4_perception_launch tier4_planning_launch tier4_sensing_launch tier4_simulator_launch tier4_system_launch tier4_vehicle_launch fastrtps cyclonedds lanelet2 lanelet2_core lanelet2_examples lanelet2_io lanelet2_maps lanelet2_matching lanelet2_projection lanelet2_python lanelet2_routing lanelet2_traffic_rules lanelet2_validation sophus angles behaviortree_cpp_v3 bond bond_core bondcpp bondpy smclib test_bond cudnn_cmake_module diagnostic_aggregator diagnostic_common_diagnostics diagnostic_updater diagnostics self_test filters geodesy geographic_info geographic_msgs grid_map grid_map_cmake_helpers grid_map_core grid_map_costmap_2d grid_map_cv grid_map_demos grid_map_filters grid_map_loader grid_map_msgs grid_map_octomap grid_map_pcl grid_map_ros grid_map_rviz_plugin grid_map_sdf grid_map_visualization mrt_cmake_modules nav2_amcl nav2_behavior_tree nav2_behaviors nav2_bringup nav2_bt_navigator nav2_collision_monitor nav2_common nav2_controller nav2_core nav2_costmap_2d costmap_queue dwb_core dwb_critics dwb_msgs dwb_plugins nav2_dwb_controller nav_2d_msgs nav_2d_utils nav2_lifecycle_manager nav2_map_server nav2_msgs nav2_navfn_planner nav2_planner nav2_regulated_pure_pursuit_controller nav2_rotation_shim_controller nav2_rviz_plugins nav2_simple_commander nav2_smac_planner nav2_smoother nav2_system_tests nav2_theta_star_planner nav2_util nav2_velocity_smoother nav2_voxel_grid nav2_waypoint_follower navigation2 dynamic_edt_3d octomap octovis octomap_msgs osqp_vendor pacmod3_msgs pcl_msgs pcl_conversions pcl_ros perception_pcl point_cloud_msg_wrapper radar_msgs can_msgs rqt_tf_tree tensorrt_cmake_module topic_tools topic_tools_interfaces tvm_vendor cv_bridge image_geometry opencv_tests vision_opencv xacro rviz2 rviz_assimp_vendor rviz_common rviz_default_plugins rviz_ogre_vendor rviz_rendering rviz_rendering_tests rviz_visual_testing_framework dummy_perception_publisher fault_injection simple_planning_simulator image_view v4l2_camera aion_interface carla_vehicle_can cgi430_can_driver cgi610_driver ARS408_driver common_sensor_launch kunyi_sensor_kit_description kunyi_sensor_kit_launch classformsg node_v2x data_format_dump data_preprocess_launch lidar_centerpoint_collect lidar_saver message_sync time_cal camera_calibration direct_visual_lidar_calibration multi_lidar_calibration

ROS Distro
github

API Docs
Browse Code

Package Summary

Tags No category tags.
Version 0.0.1
License Apache License 2.0
Build type AMENT_CMAKE
Use RECOMMENDED

Repository Summary

Description
Checkout URI https://github.com/ieiauto/autodrrt.git
VCS Type git
VCS Version main
Last Updated 2025-05-30
Dev Status UNKNOWN
Released UNRELEASED
Tags No category tags.
Contributing Help Wanted (-)
Good First Issues (-)
Pull Requests to Review (-)

Package Description

tensorrt library implementation for yolox

Additional Links

No additional links.

Maintainers

  • Daisuke Nishimatsu
  • Dan Umeda
  • Manato Hirabayashi

Authors

  • Daisuke Nishimatsu
src/autodrrt.application/perception/autoware.perception/tensorrt_yolox/README.md

tensorrt_yolox

Purpose

This package detects target objects e.g., cars, trucks, bicycles, and pedestrians on a image based on YOLOX model.

Inner-workings / Algorithms

Cite

Zheng Ge, Songtao Liu, Feng Wang, Zeming Li, Jian Sun, “YOLOX: Exceeding YOLO Series in 2021”, arXiv preprint arXiv:2107.08430, 2021 [ref]

Inputs / Outputs

Input

Name Type Description
in/image sensor_msgs/Image The input image

Output

Name Type Description
out/objects tier4_perception_msgs/DetectedObjectsWithFeature The detected objects with 2D bounding boxes
out/image sensor_msgs/Image The image with 2D bounding boxes for visualization

Parameters

Core Parameters

Name Type Default Value Description
score_threshold float 0.3 If the objectness score is less than this value, the object is ignored in yolox layer.
nms_threshold float 0.7 The IoU threshold for NMS method

NOTE: These two parameters are only valid for “plain” model (described later).

Node Parameters

Name Type Default Value Description
model_path string ”” The onnx file name for yolox model
label_path string ”” The label file with label names for detected objects written on it
precision string “fp16” The inference mode: “fp32”, “fp16”, “int8”
build_only bool false shutdown node after TensorRT engine file is built
calibration_algorithm string “MinMax” Calibration algorithm to be used for quantization when precision==int8. Valid value is one of: Entropy”,(“Legacy” | “Percentile”), “MinMax”]
dla_core_id int -1 If positive ID value is specified, the node assign inference task to the DLA core
quantize_first_layer bool false If true, set the operating precision for the first (input) layer to be fp16. This option is valid only when precision==int8
quantize_last_layer bool false If true, set the operating precision for the last (output) layer to be fp16. This option is valid only when precision==int8
profile_per_layer bool false If true, profiler function will be enabled. Since the profile function may affect execution speed, it is recommended to set this flag true only for development purpose.
clip_value double 0.0 If positive value is specified, the value of each layer output will be clipped between [0.0, clip_value]. This option is valid only when precision==int8 and used to manually specify the dynamic range instead of using any calibration
preprocess_on_gpu bool true If true, pre-processing is performed on GPU
calibration_image_list_path string ”” Path to a file which contains path to images. Those images will be used for int8 quantization.

Assumptions / Known limits

The label contained in detected 2D bounding boxes (i.e., out/objects) will be either one of the followings:

  • CAR
  • PEDESTRIAN (“PERSON” will also be categorized as “PEDESTRIAN”)
  • BUS
  • TRUCK
  • BICYCLE
  • MOTORCYCLE

If other labels (case insensitive) are contained in the file specified via the label_file parameter, those are labeled as UNKNOWN, while detected rectangles are drawn in the visualization result (out/image).

Onnx model

A sample model (named yolox-tiny.onnx) is downloaded by ansible script on env preparation stage, if not, please, follow Manual downloading of artifacts. To accelerate Non-maximum-suppression (NMS), which is one of the common post-process after object detection inference, EfficientNMS_TRT module is attached after the ordinal YOLOX (tiny) network. The EfficientNMS_TRT module contains fixed values for score_threshold and nms_threshold in it, hence these parameters are ignored when users specify ONNX models including this module.

This package accepts both EfficientNMS_TRT attached ONNXs and models published from the official YOLOX repository (we referred to them as “plain” models).

In addition to yolox-tiny.onnx, a custom model named yolox-sPlus-opt.onnx is either available. This model is based on YOLOX-s and tuned to perform more accurate detection with almost comparable execution speed with yolox-tiny. To get better results with this model, users are recommended to use some specific running arguments such as precision:=int8, calibration_algorithm:=Entropy, clip_value:=6.0. Users can refer launch/yolox_sPlus_opt.launch.xml to see how this model can be used.

All models are automatically converted to TensorRT format. These converted files will be saved in the same directory as specified ONNX files with .engine filename extension and reused from the next run. The conversion process may take a while (typically 10 to 20 minutes) and the inference process is blocked until complete the conversion, so it will take some time until detection results are published (even until appearing in the topic list) on the first run

Package acceptable model generation

To convert users’ own model that saved in PyTorch’s pth format into ONNX, users can exploit the converter offered by the official repository. For the convenience, only procedures are described below. Please refer the official document for more detail.

File truncated at 100 lines see the full file

CHANGELOG
No CHANGELOG found.

Launch files

  • launch/yolox.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_s_plus_opt.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_tiny.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-tiny]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: fp16]
      • calibration_algorithm [default: MinMax]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 0.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged tensorrt_yolox at Robotics Stack Exchange

No version for distro galactic showing github. Known supported distros are highlighted in the buttons above.
Package symbol

tensorrt_yolox package from autodrrt repo

autonomous_emergency_braking control_performance_analysis control_validator external_cmd_selector joy_controller lane_departure_checker mpc_lateral_controller obstacle_collision_checker operation_mode_transition_manager pid_longitudinal_controller predicted_path_checker pure_pursuit shift_decider trajectory_follower_base trajectory_follower_node vehicle_cmd_gate diagnostic_converter kinematic_evaluator localization_evaluator planning_evaluator ekf_localizer geo_pose_projector gyro_odometer ar_tag_based_localizer landmark_manager localization_error_monitor localization_gnss localization_util ndt_scan_matcher pose2twist pose_initializer pose_instability_detector stop_filter tree_structured_parzen_estimator twist2accel yabloc_common yabloc_image_processing yabloc_monitor yabloc_particle_filter yabloc_pose_initializer map_height_fitter map_loader map_projection_loader map_tf_generator lanelet2_map_preprocessor ros2_bevdet bytetrack cluster_merger compare_map_segmentation crosswalk_traffic_light_estimator detected_object_feature_remover detected_object_validation detection_by_tracker elevation_map_loader euclidean_cluster front_vehicle_velocity_estimator ground_segmentation heatmap_visualizer image_projection_based_fusion lidar_apollo_instance_segmentation lidar_apollo_segmentation_tvm lidar_apollo_segmentation_tvm_nodes lidar_centerpoint lidar_centerpoint_tvm map_based_prediction multi_object_tracker object_merger object_range_splitter object_velocity_splitter occupancy_grid_map_outlier_filter probabilistic_occupancy_grid_map radar_crossing_objects_noise_filter radar_fusion_to_detected_object radar_object_clustering radar_object_tracker radar_tracks_msgs_converter shape_estimation simple_object_merger tensorrt_classifier tensorrt_yolo tensorrt_yolox tracking_object_merger traffic_light_arbiter traffic_light_classifier traffic_light_fine_detector traffic_light_map_based_detector traffic_light_multi_camera_fusion traffic_light_occlusion_predictor traffic_light_ssd_fine_detector traffic_light_visualization behavior_path_avoidance_by_lane_change_module behavior_path_avoidance_module behavior_path_external_request_lane_change_module behavior_path_goal_planner_module behavior_path_lane_change_module behavior_path_planner behavior_path_planner_common behavior_path_side_shift_module behavior_path_start_planner_module behavior_velocity_blind_spot_module behavior_velocity_crosswalk_module behavior_velocity_detection_area_module behavior_velocity_intersection_module behavior_velocity_no_drivable_lane_module behavior_velocity_no_stopping_area_module behavior_velocity_occlusion_spot_module behavior_velocity_out_of_lane_module behavior_velocity_planner behavior_velocity_planner_common behavior_velocity_run_out_module behavior_velocity_speed_bump_module behavior_velocity_stop_line_module behavior_velocity_template_module behavior_velocity_traffic_light_module behavior_velocity_virtual_traffic_light_module behavior_velocity_walkway_module costmap_generator external_velocity_limit_selector freespace_planner freespace_planning_algorithms mission_planner motion_velocity_smoother objects_of_interest_marker_interface obstacle_avoidance_planner obstacle_cruise_planner obstacle_stop_planner obstacle_velocity_limiter path_smoother planning_debug_tools planning_test_utils planning_topic_converter planning_validator route_handler rtc_interface rtc_replayer bezier_sampler frenet_planner path_sampler sampler_common scenario_selector static_centerline_optimizer surround_obstacle_checker gnss_poser image_diagnostics image_transport_decompressor imu_corrector livox_tag_filter pointcloud_preprocessor radar_scan_to_pointcloud2 radar_static_pointcloud_filter radar_threshold_filter radar_tracks_noise_filter tier4_pcl_extensions vehicle_velocity_converter autoware_auto_msgs_adapter bluetooth_monitor component_state_monitor default_ad_api ad_api_adaptors ad_api_visualizers automatic_pose_initializer diagnostic_graph_aggregator dummy_diag_publisher dummy_infrastructure duplicated_node_checker emergency_handler mrm_comfortable_stop_operator mrm_emergency_stop_operator system_error_monitor system_monitor topic_state_monitor velodyne_monitor accel_brake_map_calibrator external_cmd_converter raw_vehicle_cmd_converter steer_offset_estimator vehicle_info_util launch loop_sim ssh_machine gpudirect-dds autoware_ad_api_specs autoware_adapi_v1_msgs autoware_adapi_version_msgs autoware_auto_common autoware_auto_geometry autoware_auto_control_msgs autoware_auto_geometry_msgs autoware_auto_mapping_msgs autoware_auto_msgs autoware_auto_perception_msgs autoware_auto_planning_msgs autoware_auto_system_msgs autoware_auto_vehicle_msgs autoware_auto_perception_rviz_plugin autoware_auto_tf2 autoware_cmake autoware_lint_common autoware_utils lanelet2_extension autoware_common_msgs autoware_control_msgs autoware_localization_msgs autoware_map_msgs autoware_perception_msgs autoware_planning_msgs autoware_sensing_msgs autoware_system_msgs autoware_vehicle_msgs autoware_point_types autoware_testing bag_time_manager_rviz_plugin component_interface_specs component_interface_tools component_interface_utils cuda_utils fake_test_node geography_utils global_parameter_loader glog_component goal_distance_calculator grid_map_utils interpolation kalman_filter motion_utils object_recognition_utils osqp_interface path_distance_calculator perception_utils polar_grid qp_interface rtc_manager_rviz_plugin signal_processing tensorrt_common tier4_adapi_rviz_plugin tier4_api_utils tier4_automatic_goal_rviz_plugin tier4_autoware_utils tier4_calibration_rviz_plugin tier4_camera_view_rviz_plugin tier4_control_rviz_plugin tier4_datetime_rviz_plugin tier4_debug_rviz_plugin tier4_debug_tools tier4_localization_rviz_plugin tier4_perception_rviz_plugin tier4_planning_rviz_plugin tier4_screen_capture_rviz_plugin tier4_simulated_clock_rviz_plugin tier4_state_rviz_plugin tier4_system_rviz_plugin tier4_target_object_type_rviz_plugin tier4_traffic_light_rviz_plugin tier4_vehicle_rviz_plugin time_utils simulator_compatibility_test traffic_light_recognition_marker_publisher traffic_light_utils tvm_utility dma_customer_msg dma_transfer eagleye_coordinate eagleye_navigation eagleye_msgs eagleye_rt eagleye_can_velocity_converter eagleye_fix2kml eagleye_geo_pose_converter eagleye_geo_pose_fusion eagleye_gnss_converter eagleye_tf llh_converter morai_msgs mussp ndt_omp orocos_kdl python_orocos_kdl pointcloud_to_laserscan rtklib_bridge rtklib_msgs autoware_external_api_msgs autoware_iv_external_api_adaptor autoware_iv_internal_api_adaptor awapi_awiv_adapter tier4_api_msgs tier4_auto_msgs_converter tier4_control_msgs tier4_debug_msgs tier4_external_api_msgs tier4_hmi_msgs tier4_localization_msgs tier4_map_msgs tier4_perception_msgs tier4_planning_msgs tier4_rtc_msgs tier4_simulation_msgs tier4_system_msgs tier4_v2x_msgs tier4_vehicle_msgs tier4_autoware_api_launch tier4_control_launch tier4_localization_launch tier4_map_launch tier4_perception_launch tier4_planning_launch tier4_sensing_launch tier4_simulator_launch tier4_system_launch tier4_vehicle_launch fastrtps cyclonedds lanelet2 lanelet2_core lanelet2_examples lanelet2_io lanelet2_maps lanelet2_matching lanelet2_projection lanelet2_python lanelet2_routing lanelet2_traffic_rules lanelet2_validation sophus angles behaviortree_cpp_v3 bond bond_core bondcpp bondpy smclib test_bond cudnn_cmake_module diagnostic_aggregator diagnostic_common_diagnostics diagnostic_updater diagnostics self_test filters geodesy geographic_info geographic_msgs grid_map grid_map_cmake_helpers grid_map_core grid_map_costmap_2d grid_map_cv grid_map_demos grid_map_filters grid_map_loader grid_map_msgs grid_map_octomap grid_map_pcl grid_map_ros grid_map_rviz_plugin grid_map_sdf grid_map_visualization mrt_cmake_modules nav2_amcl nav2_behavior_tree nav2_behaviors nav2_bringup nav2_bt_navigator nav2_collision_monitor nav2_common nav2_controller nav2_core nav2_costmap_2d costmap_queue dwb_core dwb_critics dwb_msgs dwb_plugins nav2_dwb_controller nav_2d_msgs nav_2d_utils nav2_lifecycle_manager nav2_map_server nav2_msgs nav2_navfn_planner nav2_planner nav2_regulated_pure_pursuit_controller nav2_rotation_shim_controller nav2_rviz_plugins nav2_simple_commander nav2_smac_planner nav2_smoother nav2_system_tests nav2_theta_star_planner nav2_util nav2_velocity_smoother nav2_voxel_grid nav2_waypoint_follower navigation2 dynamic_edt_3d octomap octovis octomap_msgs osqp_vendor pacmod3_msgs pcl_msgs pcl_conversions pcl_ros perception_pcl point_cloud_msg_wrapper radar_msgs can_msgs rqt_tf_tree tensorrt_cmake_module topic_tools topic_tools_interfaces tvm_vendor cv_bridge image_geometry opencv_tests vision_opencv xacro rviz2 rviz_assimp_vendor rviz_common rviz_default_plugins rviz_ogre_vendor rviz_rendering rviz_rendering_tests rviz_visual_testing_framework dummy_perception_publisher fault_injection simple_planning_simulator image_view v4l2_camera aion_interface carla_vehicle_can cgi430_can_driver cgi610_driver ARS408_driver common_sensor_launch kunyi_sensor_kit_description kunyi_sensor_kit_launch classformsg node_v2x data_format_dump data_preprocess_launch lidar_centerpoint_collect lidar_saver message_sync time_cal camera_calibration direct_visual_lidar_calibration multi_lidar_calibration

ROS Distro
github

API Docs
Browse Code

Package Summary

Tags No category tags.
Version 0.0.1
License Apache License 2.0
Build type AMENT_CMAKE
Use RECOMMENDED

Repository Summary

Description
Checkout URI https://github.com/ieiauto/autodrrt.git
VCS Type git
VCS Version main
Last Updated 2025-05-30
Dev Status UNKNOWN
Released UNRELEASED
Tags No category tags.
Contributing Help Wanted (-)
Good First Issues (-)
Pull Requests to Review (-)

Package Description

tensorrt library implementation for yolox

Additional Links

No additional links.

Maintainers

  • Daisuke Nishimatsu
  • Dan Umeda
  • Manato Hirabayashi

Authors

  • Daisuke Nishimatsu
src/autodrrt.application/perception/autoware.perception/tensorrt_yolox/README.md

tensorrt_yolox

Purpose

This package detects target objects e.g., cars, trucks, bicycles, and pedestrians on a image based on YOLOX model.

Inner-workings / Algorithms

Cite

Zheng Ge, Songtao Liu, Feng Wang, Zeming Li, Jian Sun, “YOLOX: Exceeding YOLO Series in 2021”, arXiv preprint arXiv:2107.08430, 2021 [ref]

Inputs / Outputs

Input

Name Type Description
in/image sensor_msgs/Image The input image

Output

Name Type Description
out/objects tier4_perception_msgs/DetectedObjectsWithFeature The detected objects with 2D bounding boxes
out/image sensor_msgs/Image The image with 2D bounding boxes for visualization

Parameters

Core Parameters

Name Type Default Value Description
score_threshold float 0.3 If the objectness score is less than this value, the object is ignored in yolox layer.
nms_threshold float 0.7 The IoU threshold for NMS method

NOTE: These two parameters are only valid for “plain” model (described later).

Node Parameters

Name Type Default Value Description
model_path string ”” The onnx file name for yolox model
label_path string ”” The label file with label names for detected objects written on it
precision string “fp16” The inference mode: “fp32”, “fp16”, “int8”
build_only bool false shutdown node after TensorRT engine file is built
calibration_algorithm string “MinMax” Calibration algorithm to be used for quantization when precision==int8. Valid value is one of: Entropy”,(“Legacy” | “Percentile”), “MinMax”]
dla_core_id int -1 If positive ID value is specified, the node assign inference task to the DLA core
quantize_first_layer bool false If true, set the operating precision for the first (input) layer to be fp16. This option is valid only when precision==int8
quantize_last_layer bool false If true, set the operating precision for the last (output) layer to be fp16. This option is valid only when precision==int8
profile_per_layer bool false If true, profiler function will be enabled. Since the profile function may affect execution speed, it is recommended to set this flag true only for development purpose.
clip_value double 0.0 If positive value is specified, the value of each layer output will be clipped between [0.0, clip_value]. This option is valid only when precision==int8 and used to manually specify the dynamic range instead of using any calibration
preprocess_on_gpu bool true If true, pre-processing is performed on GPU
calibration_image_list_path string ”” Path to a file which contains path to images. Those images will be used for int8 quantization.

Assumptions / Known limits

The label contained in detected 2D bounding boxes (i.e., out/objects) will be either one of the followings:

  • CAR
  • PEDESTRIAN (“PERSON” will also be categorized as “PEDESTRIAN”)
  • BUS
  • TRUCK
  • BICYCLE
  • MOTORCYCLE

If other labels (case insensitive) are contained in the file specified via the label_file parameter, those are labeled as UNKNOWN, while detected rectangles are drawn in the visualization result (out/image).

Onnx model

A sample model (named yolox-tiny.onnx) is downloaded by ansible script on env preparation stage, if not, please, follow Manual downloading of artifacts. To accelerate Non-maximum-suppression (NMS), which is one of the common post-process after object detection inference, EfficientNMS_TRT module is attached after the ordinal YOLOX (tiny) network. The EfficientNMS_TRT module contains fixed values for score_threshold and nms_threshold in it, hence these parameters are ignored when users specify ONNX models including this module.

This package accepts both EfficientNMS_TRT attached ONNXs and models published from the official YOLOX repository (we referred to them as “plain” models).

In addition to yolox-tiny.onnx, a custom model named yolox-sPlus-opt.onnx is either available. This model is based on YOLOX-s and tuned to perform more accurate detection with almost comparable execution speed with yolox-tiny. To get better results with this model, users are recommended to use some specific running arguments such as precision:=int8, calibration_algorithm:=Entropy, clip_value:=6.0. Users can refer launch/yolox_sPlus_opt.launch.xml to see how this model can be used.

All models are automatically converted to TensorRT format. These converted files will be saved in the same directory as specified ONNX files with .engine filename extension and reused from the next run. The conversion process may take a while (typically 10 to 20 minutes) and the inference process is blocked until complete the conversion, so it will take some time until detection results are published (even until appearing in the topic list) on the first run

Package acceptable model generation

To convert users’ own model that saved in PyTorch’s pth format into ONNX, users can exploit the converter offered by the official repository. For the convenience, only procedures are described below. Please refer the official document for more detail.

File truncated at 100 lines see the full file

CHANGELOG
No CHANGELOG found.

Launch files

  • launch/yolox.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_s_plus_opt.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_tiny.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-tiny]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: fp16]
      • calibration_algorithm [default: MinMax]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 0.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged tensorrt_yolox at Robotics Stack Exchange

No version for distro iron showing github. Known supported distros are highlighted in the buttons above.
Package symbol

tensorrt_yolox package from autodrrt repo

autonomous_emergency_braking control_performance_analysis control_validator external_cmd_selector joy_controller lane_departure_checker mpc_lateral_controller obstacle_collision_checker operation_mode_transition_manager pid_longitudinal_controller predicted_path_checker pure_pursuit shift_decider trajectory_follower_base trajectory_follower_node vehicle_cmd_gate diagnostic_converter kinematic_evaluator localization_evaluator planning_evaluator ekf_localizer geo_pose_projector gyro_odometer ar_tag_based_localizer landmark_manager localization_error_monitor localization_gnss localization_util ndt_scan_matcher pose2twist pose_initializer pose_instability_detector stop_filter tree_structured_parzen_estimator twist2accel yabloc_common yabloc_image_processing yabloc_monitor yabloc_particle_filter yabloc_pose_initializer map_height_fitter map_loader map_projection_loader map_tf_generator lanelet2_map_preprocessor ros2_bevdet bytetrack cluster_merger compare_map_segmentation crosswalk_traffic_light_estimator detected_object_feature_remover detected_object_validation detection_by_tracker elevation_map_loader euclidean_cluster front_vehicle_velocity_estimator ground_segmentation heatmap_visualizer image_projection_based_fusion lidar_apollo_instance_segmentation lidar_apollo_segmentation_tvm lidar_apollo_segmentation_tvm_nodes lidar_centerpoint lidar_centerpoint_tvm map_based_prediction multi_object_tracker object_merger object_range_splitter object_velocity_splitter occupancy_grid_map_outlier_filter probabilistic_occupancy_grid_map radar_crossing_objects_noise_filter radar_fusion_to_detected_object radar_object_clustering radar_object_tracker radar_tracks_msgs_converter shape_estimation simple_object_merger tensorrt_classifier tensorrt_yolo tensorrt_yolox tracking_object_merger traffic_light_arbiter traffic_light_classifier traffic_light_fine_detector traffic_light_map_based_detector traffic_light_multi_camera_fusion traffic_light_occlusion_predictor traffic_light_ssd_fine_detector traffic_light_visualization behavior_path_avoidance_by_lane_change_module behavior_path_avoidance_module behavior_path_external_request_lane_change_module behavior_path_goal_planner_module behavior_path_lane_change_module behavior_path_planner behavior_path_planner_common behavior_path_side_shift_module behavior_path_start_planner_module behavior_velocity_blind_spot_module behavior_velocity_crosswalk_module behavior_velocity_detection_area_module behavior_velocity_intersection_module behavior_velocity_no_drivable_lane_module behavior_velocity_no_stopping_area_module behavior_velocity_occlusion_spot_module behavior_velocity_out_of_lane_module behavior_velocity_planner behavior_velocity_planner_common behavior_velocity_run_out_module behavior_velocity_speed_bump_module behavior_velocity_stop_line_module behavior_velocity_template_module behavior_velocity_traffic_light_module behavior_velocity_virtual_traffic_light_module behavior_velocity_walkway_module costmap_generator external_velocity_limit_selector freespace_planner freespace_planning_algorithms mission_planner motion_velocity_smoother objects_of_interest_marker_interface obstacle_avoidance_planner obstacle_cruise_planner obstacle_stop_planner obstacle_velocity_limiter path_smoother planning_debug_tools planning_test_utils planning_topic_converter planning_validator route_handler rtc_interface rtc_replayer bezier_sampler frenet_planner path_sampler sampler_common scenario_selector static_centerline_optimizer surround_obstacle_checker gnss_poser image_diagnostics image_transport_decompressor imu_corrector livox_tag_filter pointcloud_preprocessor radar_scan_to_pointcloud2 radar_static_pointcloud_filter radar_threshold_filter radar_tracks_noise_filter tier4_pcl_extensions vehicle_velocity_converter autoware_auto_msgs_adapter bluetooth_monitor component_state_monitor default_ad_api ad_api_adaptors ad_api_visualizers automatic_pose_initializer diagnostic_graph_aggregator dummy_diag_publisher dummy_infrastructure duplicated_node_checker emergency_handler mrm_comfortable_stop_operator mrm_emergency_stop_operator system_error_monitor system_monitor topic_state_monitor velodyne_monitor accel_brake_map_calibrator external_cmd_converter raw_vehicle_cmd_converter steer_offset_estimator vehicle_info_util launch loop_sim ssh_machine gpudirect-dds autoware_ad_api_specs autoware_adapi_v1_msgs autoware_adapi_version_msgs autoware_auto_common autoware_auto_geometry autoware_auto_control_msgs autoware_auto_geometry_msgs autoware_auto_mapping_msgs autoware_auto_msgs autoware_auto_perception_msgs autoware_auto_planning_msgs autoware_auto_system_msgs autoware_auto_vehicle_msgs autoware_auto_perception_rviz_plugin autoware_auto_tf2 autoware_cmake autoware_lint_common autoware_utils lanelet2_extension autoware_common_msgs autoware_control_msgs autoware_localization_msgs autoware_map_msgs autoware_perception_msgs autoware_planning_msgs autoware_sensing_msgs autoware_system_msgs autoware_vehicle_msgs autoware_point_types autoware_testing bag_time_manager_rviz_plugin component_interface_specs component_interface_tools component_interface_utils cuda_utils fake_test_node geography_utils global_parameter_loader glog_component goal_distance_calculator grid_map_utils interpolation kalman_filter motion_utils object_recognition_utils osqp_interface path_distance_calculator perception_utils polar_grid qp_interface rtc_manager_rviz_plugin signal_processing tensorrt_common tier4_adapi_rviz_plugin tier4_api_utils tier4_automatic_goal_rviz_plugin tier4_autoware_utils tier4_calibration_rviz_plugin tier4_camera_view_rviz_plugin tier4_control_rviz_plugin tier4_datetime_rviz_plugin tier4_debug_rviz_plugin tier4_debug_tools tier4_localization_rviz_plugin tier4_perception_rviz_plugin tier4_planning_rviz_plugin tier4_screen_capture_rviz_plugin tier4_simulated_clock_rviz_plugin tier4_state_rviz_plugin tier4_system_rviz_plugin tier4_target_object_type_rviz_plugin tier4_traffic_light_rviz_plugin tier4_vehicle_rviz_plugin time_utils simulator_compatibility_test traffic_light_recognition_marker_publisher traffic_light_utils tvm_utility dma_customer_msg dma_transfer eagleye_coordinate eagleye_navigation eagleye_msgs eagleye_rt eagleye_can_velocity_converter eagleye_fix2kml eagleye_geo_pose_converter eagleye_geo_pose_fusion eagleye_gnss_converter eagleye_tf llh_converter morai_msgs mussp ndt_omp orocos_kdl python_orocos_kdl pointcloud_to_laserscan rtklib_bridge rtklib_msgs autoware_external_api_msgs autoware_iv_external_api_adaptor autoware_iv_internal_api_adaptor awapi_awiv_adapter tier4_api_msgs tier4_auto_msgs_converter tier4_control_msgs tier4_debug_msgs tier4_external_api_msgs tier4_hmi_msgs tier4_localization_msgs tier4_map_msgs tier4_perception_msgs tier4_planning_msgs tier4_rtc_msgs tier4_simulation_msgs tier4_system_msgs tier4_v2x_msgs tier4_vehicle_msgs tier4_autoware_api_launch tier4_control_launch tier4_localization_launch tier4_map_launch tier4_perception_launch tier4_planning_launch tier4_sensing_launch tier4_simulator_launch tier4_system_launch tier4_vehicle_launch fastrtps cyclonedds lanelet2 lanelet2_core lanelet2_examples lanelet2_io lanelet2_maps lanelet2_matching lanelet2_projection lanelet2_python lanelet2_routing lanelet2_traffic_rules lanelet2_validation sophus angles behaviortree_cpp_v3 bond bond_core bondcpp bondpy smclib test_bond cudnn_cmake_module diagnostic_aggregator diagnostic_common_diagnostics diagnostic_updater diagnostics self_test filters geodesy geographic_info geographic_msgs grid_map grid_map_cmake_helpers grid_map_core grid_map_costmap_2d grid_map_cv grid_map_demos grid_map_filters grid_map_loader grid_map_msgs grid_map_octomap grid_map_pcl grid_map_ros grid_map_rviz_plugin grid_map_sdf grid_map_visualization mrt_cmake_modules nav2_amcl nav2_behavior_tree nav2_behaviors nav2_bringup nav2_bt_navigator nav2_collision_monitor nav2_common nav2_controller nav2_core nav2_costmap_2d costmap_queue dwb_core dwb_critics dwb_msgs dwb_plugins nav2_dwb_controller nav_2d_msgs nav_2d_utils nav2_lifecycle_manager nav2_map_server nav2_msgs nav2_navfn_planner nav2_planner nav2_regulated_pure_pursuit_controller nav2_rotation_shim_controller nav2_rviz_plugins nav2_simple_commander nav2_smac_planner nav2_smoother nav2_system_tests nav2_theta_star_planner nav2_util nav2_velocity_smoother nav2_voxel_grid nav2_waypoint_follower navigation2 dynamic_edt_3d octomap octovis octomap_msgs osqp_vendor pacmod3_msgs pcl_msgs pcl_conversions pcl_ros perception_pcl point_cloud_msg_wrapper radar_msgs can_msgs rqt_tf_tree tensorrt_cmake_module topic_tools topic_tools_interfaces tvm_vendor cv_bridge image_geometry opencv_tests vision_opencv xacro rviz2 rviz_assimp_vendor rviz_common rviz_default_plugins rviz_ogre_vendor rviz_rendering rviz_rendering_tests rviz_visual_testing_framework dummy_perception_publisher fault_injection simple_planning_simulator image_view v4l2_camera aion_interface carla_vehicle_can cgi430_can_driver cgi610_driver ARS408_driver common_sensor_launch kunyi_sensor_kit_description kunyi_sensor_kit_launch classformsg node_v2x data_format_dump data_preprocess_launch lidar_centerpoint_collect lidar_saver message_sync time_cal camera_calibration direct_visual_lidar_calibration multi_lidar_calibration

ROS Distro
github

API Docs
Browse Code

Package Summary

Tags No category tags.
Version 0.0.1
License Apache License 2.0
Build type AMENT_CMAKE
Use RECOMMENDED

Repository Summary

Description
Checkout URI https://github.com/ieiauto/autodrrt.git
VCS Type git
VCS Version main
Last Updated 2025-05-30
Dev Status UNKNOWN
Released UNRELEASED
Tags No category tags.
Contributing Help Wanted (-)
Good First Issues (-)
Pull Requests to Review (-)

Package Description

tensorrt library implementation for yolox

Additional Links

No additional links.

Maintainers

  • Daisuke Nishimatsu
  • Dan Umeda
  • Manato Hirabayashi

Authors

  • Daisuke Nishimatsu
src/autodrrt.application/perception/autoware.perception/tensorrt_yolox/README.md

tensorrt_yolox

Purpose

This package detects target objects e.g., cars, trucks, bicycles, and pedestrians on a image based on YOLOX model.

Inner-workings / Algorithms

Cite

Zheng Ge, Songtao Liu, Feng Wang, Zeming Li, Jian Sun, “YOLOX: Exceeding YOLO Series in 2021”, arXiv preprint arXiv:2107.08430, 2021 [ref]

Inputs / Outputs

Input

Name Type Description
in/image sensor_msgs/Image The input image

Output

Name Type Description
out/objects tier4_perception_msgs/DetectedObjectsWithFeature The detected objects with 2D bounding boxes
out/image sensor_msgs/Image The image with 2D bounding boxes for visualization

Parameters

Core Parameters

Name Type Default Value Description
score_threshold float 0.3 If the objectness score is less than this value, the object is ignored in yolox layer.
nms_threshold float 0.7 The IoU threshold for NMS method

NOTE: These two parameters are only valid for “plain” model (described later).

Node Parameters

Name Type Default Value Description
model_path string ”” The onnx file name for yolox model
label_path string ”” The label file with label names for detected objects written on it
precision string “fp16” The inference mode: “fp32”, “fp16”, “int8”
build_only bool false shutdown node after TensorRT engine file is built
calibration_algorithm string “MinMax” Calibration algorithm to be used for quantization when precision==int8. Valid value is one of: Entropy”,(“Legacy” | “Percentile”), “MinMax”]
dla_core_id int -1 If positive ID value is specified, the node assign inference task to the DLA core
quantize_first_layer bool false If true, set the operating precision for the first (input) layer to be fp16. This option is valid only when precision==int8
quantize_last_layer bool false If true, set the operating precision for the last (output) layer to be fp16. This option is valid only when precision==int8
profile_per_layer bool false If true, profiler function will be enabled. Since the profile function may affect execution speed, it is recommended to set this flag true only for development purpose.
clip_value double 0.0 If positive value is specified, the value of each layer output will be clipped between [0.0, clip_value]. This option is valid only when precision==int8 and used to manually specify the dynamic range instead of using any calibration
preprocess_on_gpu bool true If true, pre-processing is performed on GPU
calibration_image_list_path string ”” Path to a file which contains path to images. Those images will be used for int8 quantization.

Assumptions / Known limits

The label contained in detected 2D bounding boxes (i.e., out/objects) will be either one of the followings:

  • CAR
  • PEDESTRIAN (“PERSON” will also be categorized as “PEDESTRIAN”)
  • BUS
  • TRUCK
  • BICYCLE
  • MOTORCYCLE

If other labels (case insensitive) are contained in the file specified via the label_file parameter, those are labeled as UNKNOWN, while detected rectangles are drawn in the visualization result (out/image).

Onnx model

A sample model (named yolox-tiny.onnx) is downloaded by ansible script on env preparation stage, if not, please, follow Manual downloading of artifacts. To accelerate Non-maximum-suppression (NMS), which is one of the common post-process after object detection inference, EfficientNMS_TRT module is attached after the ordinal YOLOX (tiny) network. The EfficientNMS_TRT module contains fixed values for score_threshold and nms_threshold in it, hence these parameters are ignored when users specify ONNX models including this module.

This package accepts both EfficientNMS_TRT attached ONNXs and models published from the official YOLOX repository (we referred to them as “plain” models).

In addition to yolox-tiny.onnx, a custom model named yolox-sPlus-opt.onnx is either available. This model is based on YOLOX-s and tuned to perform more accurate detection with almost comparable execution speed with yolox-tiny. To get better results with this model, users are recommended to use some specific running arguments such as precision:=int8, calibration_algorithm:=Entropy, clip_value:=6.0. Users can refer launch/yolox_sPlus_opt.launch.xml to see how this model can be used.

All models are automatically converted to TensorRT format. These converted files will be saved in the same directory as specified ONNX files with .engine filename extension and reused from the next run. The conversion process may take a while (typically 10 to 20 minutes) and the inference process is blocked until complete the conversion, so it will take some time until detection results are published (even until appearing in the topic list) on the first run

Package acceptable model generation

To convert users’ own model that saved in PyTorch’s pth format into ONNX, users can exploit the converter offered by the official repository. For the convenience, only procedures are described below. Please refer the official document for more detail.

File truncated at 100 lines see the full file

CHANGELOG
No CHANGELOG found.

Launch files

  • launch/yolox.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_s_plus_opt.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_tiny.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-tiny]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: fp16]
      • calibration_algorithm [default: MinMax]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 0.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged tensorrt_yolox at Robotics Stack Exchange

No version for distro melodic showing github. Known supported distros are highlighted in the buttons above.
Package symbol

tensorrt_yolox package from autodrrt repo

autonomous_emergency_braking control_performance_analysis control_validator external_cmd_selector joy_controller lane_departure_checker mpc_lateral_controller obstacle_collision_checker operation_mode_transition_manager pid_longitudinal_controller predicted_path_checker pure_pursuit shift_decider trajectory_follower_base trajectory_follower_node vehicle_cmd_gate diagnostic_converter kinematic_evaluator localization_evaluator planning_evaluator ekf_localizer geo_pose_projector gyro_odometer ar_tag_based_localizer landmark_manager localization_error_monitor localization_gnss localization_util ndt_scan_matcher pose2twist pose_initializer pose_instability_detector stop_filter tree_structured_parzen_estimator twist2accel yabloc_common yabloc_image_processing yabloc_monitor yabloc_particle_filter yabloc_pose_initializer map_height_fitter map_loader map_projection_loader map_tf_generator lanelet2_map_preprocessor ros2_bevdet bytetrack cluster_merger compare_map_segmentation crosswalk_traffic_light_estimator detected_object_feature_remover detected_object_validation detection_by_tracker elevation_map_loader euclidean_cluster front_vehicle_velocity_estimator ground_segmentation heatmap_visualizer image_projection_based_fusion lidar_apollo_instance_segmentation lidar_apollo_segmentation_tvm lidar_apollo_segmentation_tvm_nodes lidar_centerpoint lidar_centerpoint_tvm map_based_prediction multi_object_tracker object_merger object_range_splitter object_velocity_splitter occupancy_grid_map_outlier_filter probabilistic_occupancy_grid_map radar_crossing_objects_noise_filter radar_fusion_to_detected_object radar_object_clustering radar_object_tracker radar_tracks_msgs_converter shape_estimation simple_object_merger tensorrt_classifier tensorrt_yolo tensorrt_yolox tracking_object_merger traffic_light_arbiter traffic_light_classifier traffic_light_fine_detector traffic_light_map_based_detector traffic_light_multi_camera_fusion traffic_light_occlusion_predictor traffic_light_ssd_fine_detector traffic_light_visualization behavior_path_avoidance_by_lane_change_module behavior_path_avoidance_module behavior_path_external_request_lane_change_module behavior_path_goal_planner_module behavior_path_lane_change_module behavior_path_planner behavior_path_planner_common behavior_path_side_shift_module behavior_path_start_planner_module behavior_velocity_blind_spot_module behavior_velocity_crosswalk_module behavior_velocity_detection_area_module behavior_velocity_intersection_module behavior_velocity_no_drivable_lane_module behavior_velocity_no_stopping_area_module behavior_velocity_occlusion_spot_module behavior_velocity_out_of_lane_module behavior_velocity_planner behavior_velocity_planner_common behavior_velocity_run_out_module behavior_velocity_speed_bump_module behavior_velocity_stop_line_module behavior_velocity_template_module behavior_velocity_traffic_light_module behavior_velocity_virtual_traffic_light_module behavior_velocity_walkway_module costmap_generator external_velocity_limit_selector freespace_planner freespace_planning_algorithms mission_planner motion_velocity_smoother objects_of_interest_marker_interface obstacle_avoidance_planner obstacle_cruise_planner obstacle_stop_planner obstacle_velocity_limiter path_smoother planning_debug_tools planning_test_utils planning_topic_converter planning_validator route_handler rtc_interface rtc_replayer bezier_sampler frenet_planner path_sampler sampler_common scenario_selector static_centerline_optimizer surround_obstacle_checker gnss_poser image_diagnostics image_transport_decompressor imu_corrector livox_tag_filter pointcloud_preprocessor radar_scan_to_pointcloud2 radar_static_pointcloud_filter radar_threshold_filter radar_tracks_noise_filter tier4_pcl_extensions vehicle_velocity_converter autoware_auto_msgs_adapter bluetooth_monitor component_state_monitor default_ad_api ad_api_adaptors ad_api_visualizers automatic_pose_initializer diagnostic_graph_aggregator dummy_diag_publisher dummy_infrastructure duplicated_node_checker emergency_handler mrm_comfortable_stop_operator mrm_emergency_stop_operator system_error_monitor system_monitor topic_state_monitor velodyne_monitor accel_brake_map_calibrator external_cmd_converter raw_vehicle_cmd_converter steer_offset_estimator vehicle_info_util launch loop_sim ssh_machine gpudirect-dds autoware_ad_api_specs autoware_adapi_v1_msgs autoware_adapi_version_msgs autoware_auto_common autoware_auto_geometry autoware_auto_control_msgs autoware_auto_geometry_msgs autoware_auto_mapping_msgs autoware_auto_msgs autoware_auto_perception_msgs autoware_auto_planning_msgs autoware_auto_system_msgs autoware_auto_vehicle_msgs autoware_auto_perception_rviz_plugin autoware_auto_tf2 autoware_cmake autoware_lint_common autoware_utils lanelet2_extension autoware_common_msgs autoware_control_msgs autoware_localization_msgs autoware_map_msgs autoware_perception_msgs autoware_planning_msgs autoware_sensing_msgs autoware_system_msgs autoware_vehicle_msgs autoware_point_types autoware_testing bag_time_manager_rviz_plugin component_interface_specs component_interface_tools component_interface_utils cuda_utils fake_test_node geography_utils global_parameter_loader glog_component goal_distance_calculator grid_map_utils interpolation kalman_filter motion_utils object_recognition_utils osqp_interface path_distance_calculator perception_utils polar_grid qp_interface rtc_manager_rviz_plugin signal_processing tensorrt_common tier4_adapi_rviz_plugin tier4_api_utils tier4_automatic_goal_rviz_plugin tier4_autoware_utils tier4_calibration_rviz_plugin tier4_camera_view_rviz_plugin tier4_control_rviz_plugin tier4_datetime_rviz_plugin tier4_debug_rviz_plugin tier4_debug_tools tier4_localization_rviz_plugin tier4_perception_rviz_plugin tier4_planning_rviz_plugin tier4_screen_capture_rviz_plugin tier4_simulated_clock_rviz_plugin tier4_state_rviz_plugin tier4_system_rviz_plugin tier4_target_object_type_rviz_plugin tier4_traffic_light_rviz_plugin tier4_vehicle_rviz_plugin time_utils simulator_compatibility_test traffic_light_recognition_marker_publisher traffic_light_utils tvm_utility dma_customer_msg dma_transfer eagleye_coordinate eagleye_navigation eagleye_msgs eagleye_rt eagleye_can_velocity_converter eagleye_fix2kml eagleye_geo_pose_converter eagleye_geo_pose_fusion eagleye_gnss_converter eagleye_tf llh_converter morai_msgs mussp ndt_omp orocos_kdl python_orocos_kdl pointcloud_to_laserscan rtklib_bridge rtklib_msgs autoware_external_api_msgs autoware_iv_external_api_adaptor autoware_iv_internal_api_adaptor awapi_awiv_adapter tier4_api_msgs tier4_auto_msgs_converter tier4_control_msgs tier4_debug_msgs tier4_external_api_msgs tier4_hmi_msgs tier4_localization_msgs tier4_map_msgs tier4_perception_msgs tier4_planning_msgs tier4_rtc_msgs tier4_simulation_msgs tier4_system_msgs tier4_v2x_msgs tier4_vehicle_msgs tier4_autoware_api_launch tier4_control_launch tier4_localization_launch tier4_map_launch tier4_perception_launch tier4_planning_launch tier4_sensing_launch tier4_simulator_launch tier4_system_launch tier4_vehicle_launch fastrtps cyclonedds lanelet2 lanelet2_core lanelet2_examples lanelet2_io lanelet2_maps lanelet2_matching lanelet2_projection lanelet2_python lanelet2_routing lanelet2_traffic_rules lanelet2_validation sophus angles behaviortree_cpp_v3 bond bond_core bondcpp bondpy smclib test_bond cudnn_cmake_module diagnostic_aggregator diagnostic_common_diagnostics diagnostic_updater diagnostics self_test filters geodesy geographic_info geographic_msgs grid_map grid_map_cmake_helpers grid_map_core grid_map_costmap_2d grid_map_cv grid_map_demos grid_map_filters grid_map_loader grid_map_msgs grid_map_octomap grid_map_pcl grid_map_ros grid_map_rviz_plugin grid_map_sdf grid_map_visualization mrt_cmake_modules nav2_amcl nav2_behavior_tree nav2_behaviors nav2_bringup nav2_bt_navigator nav2_collision_monitor nav2_common nav2_controller nav2_core nav2_costmap_2d costmap_queue dwb_core dwb_critics dwb_msgs dwb_plugins nav2_dwb_controller nav_2d_msgs nav_2d_utils nav2_lifecycle_manager nav2_map_server nav2_msgs nav2_navfn_planner nav2_planner nav2_regulated_pure_pursuit_controller nav2_rotation_shim_controller nav2_rviz_plugins nav2_simple_commander nav2_smac_planner nav2_smoother nav2_system_tests nav2_theta_star_planner nav2_util nav2_velocity_smoother nav2_voxel_grid nav2_waypoint_follower navigation2 dynamic_edt_3d octomap octovis octomap_msgs osqp_vendor pacmod3_msgs pcl_msgs pcl_conversions pcl_ros perception_pcl point_cloud_msg_wrapper radar_msgs can_msgs rqt_tf_tree tensorrt_cmake_module topic_tools topic_tools_interfaces tvm_vendor cv_bridge image_geometry opencv_tests vision_opencv xacro rviz2 rviz_assimp_vendor rviz_common rviz_default_plugins rviz_ogre_vendor rviz_rendering rviz_rendering_tests rviz_visual_testing_framework dummy_perception_publisher fault_injection simple_planning_simulator image_view v4l2_camera aion_interface carla_vehicle_can cgi430_can_driver cgi610_driver ARS408_driver common_sensor_launch kunyi_sensor_kit_description kunyi_sensor_kit_launch classformsg node_v2x data_format_dump data_preprocess_launch lidar_centerpoint_collect lidar_saver message_sync time_cal camera_calibration direct_visual_lidar_calibration multi_lidar_calibration

ROS Distro
github

API Docs
Browse Code
Wiki

Package Summary

Tags No category tags.
Version 0.0.1
License Apache License 2.0
Build type AMENT_CMAKE
Use RECOMMENDED

Repository Summary

Description
Checkout URI https://github.com/ieiauto/autodrrt.git
VCS Type git
VCS Version main
Last Updated 2025-05-30
Dev Status UNKNOWN
Released UNRELEASED
Tags No category tags.
Contributing Help Wanted (-)
Good First Issues (-)
Pull Requests to Review (-)

Package Description

tensorrt library implementation for yolox

Additional Links

No additional links.

Maintainers

  • Daisuke Nishimatsu
  • Dan Umeda
  • Manato Hirabayashi

Authors

  • Daisuke Nishimatsu
src/autodrrt.application/perception/autoware.perception/tensorrt_yolox/README.md

tensorrt_yolox

Purpose

This package detects target objects e.g., cars, trucks, bicycles, and pedestrians on a image based on YOLOX model.

Inner-workings / Algorithms

Cite

Zheng Ge, Songtao Liu, Feng Wang, Zeming Li, Jian Sun, “YOLOX: Exceeding YOLO Series in 2021”, arXiv preprint arXiv:2107.08430, 2021 [ref]

Inputs / Outputs

Input

Name Type Description
in/image sensor_msgs/Image The input image

Output

Name Type Description
out/objects tier4_perception_msgs/DetectedObjectsWithFeature The detected objects with 2D bounding boxes
out/image sensor_msgs/Image The image with 2D bounding boxes for visualization

Parameters

Core Parameters

Name Type Default Value Description
score_threshold float 0.3 If the objectness score is less than this value, the object is ignored in yolox layer.
nms_threshold float 0.7 The IoU threshold for NMS method

NOTE: These two parameters are only valid for “plain” model (described later).

Node Parameters

Name Type Default Value Description
model_path string ”” The onnx file name for yolox model
label_path string ”” The label file with label names for detected objects written on it
precision string “fp16” The inference mode: “fp32”, “fp16”, “int8”
build_only bool false shutdown node after TensorRT engine file is built
calibration_algorithm string “MinMax” Calibration algorithm to be used for quantization when precision==int8. Valid value is one of: Entropy”,(“Legacy” | “Percentile”), “MinMax”]
dla_core_id int -1 If positive ID value is specified, the node assign inference task to the DLA core
quantize_first_layer bool false If true, set the operating precision for the first (input) layer to be fp16. This option is valid only when precision==int8
quantize_last_layer bool false If true, set the operating precision for the last (output) layer to be fp16. This option is valid only when precision==int8
profile_per_layer bool false If true, profiler function will be enabled. Since the profile function may affect execution speed, it is recommended to set this flag true only for development purpose.
clip_value double 0.0 If positive value is specified, the value of each layer output will be clipped between [0.0, clip_value]. This option is valid only when precision==int8 and used to manually specify the dynamic range instead of using any calibration
preprocess_on_gpu bool true If true, pre-processing is performed on GPU
calibration_image_list_path string ”” Path to a file which contains path to images. Those images will be used for int8 quantization.

Assumptions / Known limits

The label contained in detected 2D bounding boxes (i.e., out/objects) will be either one of the followings:

  • CAR
  • PEDESTRIAN (“PERSON” will also be categorized as “PEDESTRIAN”)
  • BUS
  • TRUCK
  • BICYCLE
  • MOTORCYCLE

If other labels (case insensitive) are contained in the file specified via the label_file parameter, those are labeled as UNKNOWN, while detected rectangles are drawn in the visualization result (out/image).

Onnx model

A sample model (named yolox-tiny.onnx) is downloaded by ansible script on env preparation stage, if not, please, follow Manual downloading of artifacts. To accelerate Non-maximum-suppression (NMS), which is one of the common post-process after object detection inference, EfficientNMS_TRT module is attached after the ordinal YOLOX (tiny) network. The EfficientNMS_TRT module contains fixed values for score_threshold and nms_threshold in it, hence these parameters are ignored when users specify ONNX models including this module.

This package accepts both EfficientNMS_TRT attached ONNXs and models published from the official YOLOX repository (we referred to them as “plain” models).

In addition to yolox-tiny.onnx, a custom model named yolox-sPlus-opt.onnx is either available. This model is based on YOLOX-s and tuned to perform more accurate detection with almost comparable execution speed with yolox-tiny. To get better results with this model, users are recommended to use some specific running arguments such as precision:=int8, calibration_algorithm:=Entropy, clip_value:=6.0. Users can refer launch/yolox_sPlus_opt.launch.xml to see how this model can be used.

All models are automatically converted to TensorRT format. These converted files will be saved in the same directory as specified ONNX files with .engine filename extension and reused from the next run. The conversion process may take a while (typically 10 to 20 minutes) and the inference process is blocked until complete the conversion, so it will take some time until detection results are published (even until appearing in the topic list) on the first run

Package acceptable model generation

To convert users’ own model that saved in PyTorch’s pth format into ONNX, users can exploit the converter offered by the official repository. For the convenience, only procedures are described below. Please refer the official document for more detail.

File truncated at 100 lines see the full file

CHANGELOG
No CHANGELOG found.

Launch files

  • launch/yolox.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_s_plus_opt.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_tiny.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-tiny]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: fp16]
      • calibration_algorithm [default: MinMax]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 0.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged tensorrt_yolox at Robotics Stack Exchange

No version for distro noetic showing github. Known supported distros are highlighted in the buttons above.
Package symbol

tensorrt_yolox package from autodrrt repo

autonomous_emergency_braking control_performance_analysis control_validator external_cmd_selector joy_controller lane_departure_checker mpc_lateral_controller obstacle_collision_checker operation_mode_transition_manager pid_longitudinal_controller predicted_path_checker pure_pursuit shift_decider trajectory_follower_base trajectory_follower_node vehicle_cmd_gate diagnostic_converter kinematic_evaluator localization_evaluator planning_evaluator ekf_localizer geo_pose_projector gyro_odometer ar_tag_based_localizer landmark_manager localization_error_monitor localization_gnss localization_util ndt_scan_matcher pose2twist pose_initializer pose_instability_detector stop_filter tree_structured_parzen_estimator twist2accel yabloc_common yabloc_image_processing yabloc_monitor yabloc_particle_filter yabloc_pose_initializer map_height_fitter map_loader map_projection_loader map_tf_generator lanelet2_map_preprocessor ros2_bevdet bytetrack cluster_merger compare_map_segmentation crosswalk_traffic_light_estimator detected_object_feature_remover detected_object_validation detection_by_tracker elevation_map_loader euclidean_cluster front_vehicle_velocity_estimator ground_segmentation heatmap_visualizer image_projection_based_fusion lidar_apollo_instance_segmentation lidar_apollo_segmentation_tvm lidar_apollo_segmentation_tvm_nodes lidar_centerpoint lidar_centerpoint_tvm map_based_prediction multi_object_tracker object_merger object_range_splitter object_velocity_splitter occupancy_grid_map_outlier_filter probabilistic_occupancy_grid_map radar_crossing_objects_noise_filter radar_fusion_to_detected_object radar_object_clustering radar_object_tracker radar_tracks_msgs_converter shape_estimation simple_object_merger tensorrt_classifier tensorrt_yolo tensorrt_yolox tracking_object_merger traffic_light_arbiter traffic_light_classifier traffic_light_fine_detector traffic_light_map_based_detector traffic_light_multi_camera_fusion traffic_light_occlusion_predictor traffic_light_ssd_fine_detector traffic_light_visualization behavior_path_avoidance_by_lane_change_module behavior_path_avoidance_module behavior_path_external_request_lane_change_module behavior_path_goal_planner_module behavior_path_lane_change_module behavior_path_planner behavior_path_planner_common behavior_path_side_shift_module behavior_path_start_planner_module behavior_velocity_blind_spot_module behavior_velocity_crosswalk_module behavior_velocity_detection_area_module behavior_velocity_intersection_module behavior_velocity_no_drivable_lane_module behavior_velocity_no_stopping_area_module behavior_velocity_occlusion_spot_module behavior_velocity_out_of_lane_module behavior_velocity_planner behavior_velocity_planner_common behavior_velocity_run_out_module behavior_velocity_speed_bump_module behavior_velocity_stop_line_module behavior_velocity_template_module behavior_velocity_traffic_light_module behavior_velocity_virtual_traffic_light_module behavior_velocity_walkway_module costmap_generator external_velocity_limit_selector freespace_planner freespace_planning_algorithms mission_planner motion_velocity_smoother objects_of_interest_marker_interface obstacle_avoidance_planner obstacle_cruise_planner obstacle_stop_planner obstacle_velocity_limiter path_smoother planning_debug_tools planning_test_utils planning_topic_converter planning_validator route_handler rtc_interface rtc_replayer bezier_sampler frenet_planner path_sampler sampler_common scenario_selector static_centerline_optimizer surround_obstacle_checker gnss_poser image_diagnostics image_transport_decompressor imu_corrector livox_tag_filter pointcloud_preprocessor radar_scan_to_pointcloud2 radar_static_pointcloud_filter radar_threshold_filter radar_tracks_noise_filter tier4_pcl_extensions vehicle_velocity_converter autoware_auto_msgs_adapter bluetooth_monitor component_state_monitor default_ad_api ad_api_adaptors ad_api_visualizers automatic_pose_initializer diagnostic_graph_aggregator dummy_diag_publisher dummy_infrastructure duplicated_node_checker emergency_handler mrm_comfortable_stop_operator mrm_emergency_stop_operator system_error_monitor system_monitor topic_state_monitor velodyne_monitor accel_brake_map_calibrator external_cmd_converter raw_vehicle_cmd_converter steer_offset_estimator vehicle_info_util launch loop_sim ssh_machine gpudirect-dds autoware_ad_api_specs autoware_adapi_v1_msgs autoware_adapi_version_msgs autoware_auto_common autoware_auto_geometry autoware_auto_control_msgs autoware_auto_geometry_msgs autoware_auto_mapping_msgs autoware_auto_msgs autoware_auto_perception_msgs autoware_auto_planning_msgs autoware_auto_system_msgs autoware_auto_vehicle_msgs autoware_auto_perception_rviz_plugin autoware_auto_tf2 autoware_cmake autoware_lint_common autoware_utils lanelet2_extension autoware_common_msgs autoware_control_msgs autoware_localization_msgs autoware_map_msgs autoware_perception_msgs autoware_planning_msgs autoware_sensing_msgs autoware_system_msgs autoware_vehicle_msgs autoware_point_types autoware_testing bag_time_manager_rviz_plugin component_interface_specs component_interface_tools component_interface_utils cuda_utils fake_test_node geography_utils global_parameter_loader glog_component goal_distance_calculator grid_map_utils interpolation kalman_filter motion_utils object_recognition_utils osqp_interface path_distance_calculator perception_utils polar_grid qp_interface rtc_manager_rviz_plugin signal_processing tensorrt_common tier4_adapi_rviz_plugin tier4_api_utils tier4_automatic_goal_rviz_plugin tier4_autoware_utils tier4_calibration_rviz_plugin tier4_camera_view_rviz_plugin tier4_control_rviz_plugin tier4_datetime_rviz_plugin tier4_debug_rviz_plugin tier4_debug_tools tier4_localization_rviz_plugin tier4_perception_rviz_plugin tier4_planning_rviz_plugin tier4_screen_capture_rviz_plugin tier4_simulated_clock_rviz_plugin tier4_state_rviz_plugin tier4_system_rviz_plugin tier4_target_object_type_rviz_plugin tier4_traffic_light_rviz_plugin tier4_vehicle_rviz_plugin time_utils simulator_compatibility_test traffic_light_recognition_marker_publisher traffic_light_utils tvm_utility dma_customer_msg dma_transfer eagleye_coordinate eagleye_navigation eagleye_msgs eagleye_rt eagleye_can_velocity_converter eagleye_fix2kml eagleye_geo_pose_converter eagleye_geo_pose_fusion eagleye_gnss_converter eagleye_tf llh_converter morai_msgs mussp ndt_omp orocos_kdl python_orocos_kdl pointcloud_to_laserscan rtklib_bridge rtklib_msgs autoware_external_api_msgs autoware_iv_external_api_adaptor autoware_iv_internal_api_adaptor awapi_awiv_adapter tier4_api_msgs tier4_auto_msgs_converter tier4_control_msgs tier4_debug_msgs tier4_external_api_msgs tier4_hmi_msgs tier4_localization_msgs tier4_map_msgs tier4_perception_msgs tier4_planning_msgs tier4_rtc_msgs tier4_simulation_msgs tier4_system_msgs tier4_v2x_msgs tier4_vehicle_msgs tier4_autoware_api_launch tier4_control_launch tier4_localization_launch tier4_map_launch tier4_perception_launch tier4_planning_launch tier4_sensing_launch tier4_simulator_launch tier4_system_launch tier4_vehicle_launch fastrtps cyclonedds lanelet2 lanelet2_core lanelet2_examples lanelet2_io lanelet2_maps lanelet2_matching lanelet2_projection lanelet2_python lanelet2_routing lanelet2_traffic_rules lanelet2_validation sophus angles behaviortree_cpp_v3 bond bond_core bondcpp bondpy smclib test_bond cudnn_cmake_module diagnostic_aggregator diagnostic_common_diagnostics diagnostic_updater diagnostics self_test filters geodesy geographic_info geographic_msgs grid_map grid_map_cmake_helpers grid_map_core grid_map_costmap_2d grid_map_cv grid_map_demos grid_map_filters grid_map_loader grid_map_msgs grid_map_octomap grid_map_pcl grid_map_ros grid_map_rviz_plugin grid_map_sdf grid_map_visualization mrt_cmake_modules nav2_amcl nav2_behavior_tree nav2_behaviors nav2_bringup nav2_bt_navigator nav2_collision_monitor nav2_common nav2_controller nav2_core nav2_costmap_2d costmap_queue dwb_core dwb_critics dwb_msgs dwb_plugins nav2_dwb_controller nav_2d_msgs nav_2d_utils nav2_lifecycle_manager nav2_map_server nav2_msgs nav2_navfn_planner nav2_planner nav2_regulated_pure_pursuit_controller nav2_rotation_shim_controller nav2_rviz_plugins nav2_simple_commander nav2_smac_planner nav2_smoother nav2_system_tests nav2_theta_star_planner nav2_util nav2_velocity_smoother nav2_voxel_grid nav2_waypoint_follower navigation2 dynamic_edt_3d octomap octovis octomap_msgs osqp_vendor pacmod3_msgs pcl_msgs pcl_conversions pcl_ros perception_pcl point_cloud_msg_wrapper radar_msgs can_msgs rqt_tf_tree tensorrt_cmake_module topic_tools topic_tools_interfaces tvm_vendor cv_bridge image_geometry opencv_tests vision_opencv xacro rviz2 rviz_assimp_vendor rviz_common rviz_default_plugins rviz_ogre_vendor rviz_rendering rviz_rendering_tests rviz_visual_testing_framework dummy_perception_publisher fault_injection simple_planning_simulator image_view v4l2_camera aion_interface carla_vehicle_can cgi430_can_driver cgi610_driver ARS408_driver common_sensor_launch kunyi_sensor_kit_description kunyi_sensor_kit_launch classformsg node_v2x data_format_dump data_preprocess_launch lidar_centerpoint_collect lidar_saver message_sync time_cal camera_calibration direct_visual_lidar_calibration multi_lidar_calibration

ROS Distro
github

API Docs
Browse Code
Wiki

Package Summary

Tags No category tags.
Version 0.0.1
License Apache License 2.0
Build type AMENT_CMAKE
Use RECOMMENDED

Repository Summary

Description
Checkout URI https://github.com/ieiauto/autodrrt.git
VCS Type git
VCS Version main
Last Updated 2025-05-30
Dev Status UNKNOWN
Released UNRELEASED
Tags No category tags.
Contributing Help Wanted (-)
Good First Issues (-)
Pull Requests to Review (-)

Package Description

tensorrt library implementation for yolox

Additional Links

No additional links.

Maintainers

  • Daisuke Nishimatsu
  • Dan Umeda
  • Manato Hirabayashi

Authors

  • Daisuke Nishimatsu
src/autodrrt.application/perception/autoware.perception/tensorrt_yolox/README.md

tensorrt_yolox

Purpose

This package detects target objects e.g., cars, trucks, bicycles, and pedestrians on a image based on YOLOX model.

Inner-workings / Algorithms

Cite

Zheng Ge, Songtao Liu, Feng Wang, Zeming Li, Jian Sun, “YOLOX: Exceeding YOLO Series in 2021”, arXiv preprint arXiv:2107.08430, 2021 [ref]

Inputs / Outputs

Input

Name Type Description
in/image sensor_msgs/Image The input image

Output

Name Type Description
out/objects tier4_perception_msgs/DetectedObjectsWithFeature The detected objects with 2D bounding boxes
out/image sensor_msgs/Image The image with 2D bounding boxes for visualization

Parameters

Core Parameters

Name Type Default Value Description
score_threshold float 0.3 If the objectness score is less than this value, the object is ignored in yolox layer.
nms_threshold float 0.7 The IoU threshold for NMS method

NOTE: These two parameters are only valid for “plain” model (described later).

Node Parameters

Name Type Default Value Description
model_path string ”” The onnx file name for yolox model
label_path string ”” The label file with label names for detected objects written on it
precision string “fp16” The inference mode: “fp32”, “fp16”, “int8”
build_only bool false shutdown node after TensorRT engine file is built
calibration_algorithm string “MinMax” Calibration algorithm to be used for quantization when precision==int8. Valid value is one of: Entropy”,(“Legacy” | “Percentile”), “MinMax”]
dla_core_id int -1 If positive ID value is specified, the node assign inference task to the DLA core
quantize_first_layer bool false If true, set the operating precision for the first (input) layer to be fp16. This option is valid only when precision==int8
quantize_last_layer bool false If true, set the operating precision for the last (output) layer to be fp16. This option is valid only when precision==int8
profile_per_layer bool false If true, profiler function will be enabled. Since the profile function may affect execution speed, it is recommended to set this flag true only for development purpose.
clip_value double 0.0 If positive value is specified, the value of each layer output will be clipped between [0.0, clip_value]. This option is valid only when precision==int8 and used to manually specify the dynamic range instead of using any calibration
preprocess_on_gpu bool true If true, pre-processing is performed on GPU
calibration_image_list_path string ”” Path to a file which contains path to images. Those images will be used for int8 quantization.

Assumptions / Known limits

The label contained in detected 2D bounding boxes (i.e., out/objects) will be either one of the followings:

  • CAR
  • PEDESTRIAN (“PERSON” will also be categorized as “PEDESTRIAN”)
  • BUS
  • TRUCK
  • BICYCLE
  • MOTORCYCLE

If other labels (case insensitive) are contained in the file specified via the label_file parameter, those are labeled as UNKNOWN, while detected rectangles are drawn in the visualization result (out/image).

Onnx model

A sample model (named yolox-tiny.onnx) is downloaded by ansible script on env preparation stage, if not, please, follow Manual downloading of artifacts. To accelerate Non-maximum-suppression (NMS), which is one of the common post-process after object detection inference, EfficientNMS_TRT module is attached after the ordinal YOLOX (tiny) network. The EfficientNMS_TRT module contains fixed values for score_threshold and nms_threshold in it, hence these parameters are ignored when users specify ONNX models including this module.

This package accepts both EfficientNMS_TRT attached ONNXs and models published from the official YOLOX repository (we referred to them as “plain” models).

In addition to yolox-tiny.onnx, a custom model named yolox-sPlus-opt.onnx is either available. This model is based on YOLOX-s and tuned to perform more accurate detection with almost comparable execution speed with yolox-tiny. To get better results with this model, users are recommended to use some specific running arguments such as precision:=int8, calibration_algorithm:=Entropy, clip_value:=6.0. Users can refer launch/yolox_sPlus_opt.launch.xml to see how this model can be used.

All models are automatically converted to TensorRT format. These converted files will be saved in the same directory as specified ONNX files with .engine filename extension and reused from the next run. The conversion process may take a while (typically 10 to 20 minutes) and the inference process is blocked until complete the conversion, so it will take some time until detection results are published (even until appearing in the topic list) on the first run

Package acceptable model generation

To convert users’ own model that saved in PyTorch’s pth format into ONNX, users can exploit the converter offered by the official repository. For the convenience, only procedures are described below. Please refer the official document for more detail.

File truncated at 100 lines see the full file

CHANGELOG
No CHANGELOG found.

Launch files

  • launch/yolox.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_s_plus_opt.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-sPlus-T4-960x960-pseudo-finetune]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: int8]
      • calibration_algorithm [default: Entropy]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 6.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]
  • launch/yolox_tiny.launch.xml
      • input/image [default: /sensing/camera/camera0/image_rect_color]
      • output/objects [default: /perception/object_recognition/detection/rois0]
      • model_name [default: yolox-tiny]
      • data_path [default: $(env HOME)/autoware_data]
      • model_path [default: $(var data_path)/tensorrt_yolox]
      • score_threshold [default: 0.35]
      • nms_threshold [default: 0.7]
      • precision [default: fp16]
      • calibration_algorithm [default: MinMax]
      • dla_core_id [default: -1]
      • quantize_first_layer [default: false]
      • quantize_last_layer [default: false]
      • profile_per_layer [default: false]
      • clip_value [default: 0.0]
      • preprocess_on_gpu [default: true]
      • calibration_image_list_path [default: ]
      • use_decompress [default: true]
      • build_only [default: false]

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged tensorrt_yolox at Robotics Stack Exchange