search

derek285 / yolov4_deep_sort_pytorch

MOT using deepsort and yolov4 with pytorch
MIT License
18 stars 6 forks source link

readme

Yolov4 + Deep Sort with PyTorch

remember build your own libdarknet.so and put under folder yolov4_deep_sort_pytorch/

ref : https://github.com/ZQPei/deep_sort_pytorch

Quick Start

  1. Check all dependencies installed

    pip3 install -r requirements.txt

    for user in china, you can specify pypi source to accelerate install like:

    pip3 install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple

  2. Clone this repository

    git clone https://github.com/derek285/yolov4_deep_sort_pytorch.git

  3. Download YOLOv4 parameters

    wget yolov4.weights and cfg
//并且重命名为4.cfg 和 4.weights 这个hard code有点山寨了：detector/v4detector.py ：
configPath = "4.cfg"
weightPath = "4.weights"

  4. Download deepsort parameters ckpt.t7

    cd deep_sort/deep/checkpoint
# download ckpt.t7 from
https://drive.google.com/drive/folders/1xhG0kRH1EX5B9_Iz8gQJb7UNnn_riXi6 to this folder
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.

  1. Run demo 
    
usage: python tracker.py VIDEO_PATH
                            [--help]
                            [--frame_interval FRAME_INTERVAL]
                            [--config_detection CONFIG_DETECTION]
                            [--config_deepsort CONFIG_DEEPSORT]
                            [--display]
                            [--display_width DISPLAY_WIDTH]
                            [--display_height DISPLAY_HEIGHT]
                            [--save_path SAVE_PATH]          
                            [--cpu]

yolov4 + deepsort on video file

python3 tracker.py VIDEO_PATH

yolov4 + deepsort on webcam

python3 tracker.py /dev/video0 --camera 0

todo

1. get class_id return
2. fix depends on libdarknet.so

References