This is an implement of MOT tracking algorithm deep sort. Deep sort is basicly the same with sort but added a CNN model to extract features in image of human part bounded by a detector. This CNN model is indeed a RE-ID model and the detector used in PAPER is FasterRCNN , and the original source code is HERE.
However in original code, the CNN model is implemented with tensorflow, which I'm not familier with. SO I re-implemented the CNN feature extraction model with PyTorch, and changed the CNN model a little bit. Also, I use YOLOv3 to generate bboxes instead of FasterRCNN.
- Check all dependencies installed
pip install -r requirements.txt- Download YOLOv3 parameters
cd detector/YOLOv3/weight/
wget https://pjreddie.com/media/files/yolov3.weights
wget https://pjreddie.com/media/files/yolov3-tiny.weights
cd ../../../
- Compile nms module
cd detector/YOLOv3/nms
sh build.sh
cd ../../..- Download Videos
cd data
sh download_MOT16.sh
cd ..Notice:
If compiling failed, the simplist way is to **Upgrade your pytorch >= 1.1 and torchvision >= 0.3" and you can avoid the troublesome compiling problems which are most likely caused by either gcc version too low or libraries missing.
- Run demo
usage: python yolov3_deepsort.py VIDEO_PATH
[--help]
[--frame_interval FRAME_INTERVAL]
[--config_detection CONFIG_DETECTION]
[--config_deepsort CONFIG_DEEPSORT]
[--detection_model DETECTION MODEL]
[--display]
[--display_width DISPLAY_WIDTH]
[--display_height DISPLAY_HEIGHT]
[--sample_rate SAMPLE_RATE]
[--save_path SAVE_PATH]
[--save_file SAVE_FILE]
[--cpu]
# yolov3 + deepsort
python yolov3_deepsort.py [VIDEO_NAME] --save_file [VIDEO_NAME]
Example: python yolov3_deepsort.py data/MOT16-02.avi --save_file MOT16-02
# yolov3_tiny + deepsort
python yolov3_deepsort.py [VIDEO_PATH] --config_detection ./configs/yolov3_tiny.yaml --detection_model yolov3-tiny --save_file [VIDEO_NAME]
Use --display to enable display.
Results will be saved to ./output/results.avi and ./output/results.txt.
- [Optional] Evaluate results
python -m motmetrics.apps.evaluateTracking --help
Example: python -m motmetrics.apps.evaluateTracking MOT16Labels/train output/ seq
-
paper: Simple Online and Realtime Tracking with a Deep Association Metric
-
code: nwojke/deep_sort
-
paper: YOLOv3
-
code: Joseph Redmon/yolov3