YOLOX ByteTrack-Eigen ONNX Demo

A Visual Studio project demonstrating how to perform object tracking across video frames with YOLOX, ONNX Runtime, and the ByteTrack-Eigen library.

Setup Steps:

1. Clone the repository.
2. Run the download-dependencies.bat setup file.
3. Ensure the Solution Configuration for the Visual Studio project is in Release mode.
4. Restore the NuGet packages.
5. Build the solution (Ctrl+Shift+B).
6. Place test files in the folder with the executable (sample files available below).

Example Usage:

Sample Files

• pexels-rodnae-productions-10373924.mp4
• hagrid-sample-30k-384p-yolox_tiny.onnx
• hagrid-sample-30k-384p-colormap.json
  • (To save the colormap file, right-click the link and opt for Save Link As...)

CPU Inference

YOLOXByteTrackONNXDemo.exe hagrid-sample-30k-384p-yolox_tiny.onnx pexels-rodnae-productions-10373924.mp4 hagrid-sample-30k-384p-colormap.json

DirectML Inference

YOLOXByteTrackONNXDemo.exe hagrid-sample-30k-384p-yolox_tiny.onnx pexels-rodnae-productions-10373924.mp4 hagrid-sample-30k-384p-colormap.json Dml