YOLOv9 + ByteTracker

This project integrates YOLOv9 and ByteTracker for real-time, TensorRT-optimized object detection and tracking, extending the existing TensorRT-Yolov9 implementation.

Usage

• CPP(TensorRT):

cd build/release
yolov9-bytetrack-trt.exe yolov9-c.engine test.mp4 # the video path

• Python(ONNX):

cd python
python yolov9_bytetrack.py --model yolov9-c-converted.onnx --video_path test_video.mp4 --output_video_path result.mp4

What is next?

• Python(TensorRT)

Setup

Python:

cd python
pip install -r requirement.txt

CPP:

Refer to our docs/INSTALL.md for detailed installation instructions. Note that Bytetracker is directly integrated without any modification.

Requirement

• TensorRT
• CUDA, CudaNN
• Eigen 3.3
• C++ compiler with C++17 or higher support
• CMake 3.14 or higher
• OpenCV

Acknowledgement

This project is based on the following awesome projects:

• YOLOv9 - YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information.
• ByteTrack - C++ implementation of ByteTrack algorithm.
• TensorRT-Yolov9 - C++ implementation of YOLOv9 using TensorRT API.