Using DOPE at Different Image Resolutions


The DOPE network architecture, as outlined in the original paper, can receive input images of arbitrary size and subsequently produce output belief maps of the corresponding dimensions.

However, the ONNX format used to run this network on Triton or TensorRT is not as flexible, and an ONNX-exported model does NOT support arbitrary image sizes at inference time. Instead, the desired input image dimensions must be explicitly specified when preparing the ONNX file using the script, as referenced in the quickstart.

Tutorial Walkthrough

  1. Follow steps 1-6 of the main DOPE quickstart.

  2. At step 7, run the script with the two additional arguments row and col specifying the desired input image size:

    python3 /workspaces/isaac_ros-dev/src/isaac_ros_pose_estimation/isaac_ros_dope/scripts/ --format onnx --input /tmp/models/Ketchup.pth --row 1080 --col 1920
  3. Proceed through steps 8-9.

  4. At step 10, launch the ROS 2 launch file with two additional arguments network_image_height and network_image_width specifying the desired input image size:

    ros2 launch isaac_ros_dope model_file_path:=/tmp/models/Ketchup.onnx engine_file_path:=/tmp/models/Ketchup.plan network_image_height:=1080 network_image_width:=1920 input_image_height:=1080 input_image_width:=1920
  5. Continue with the rest of the quickstart. You should now be able to detect poses in images of your desired size.