- Python3
- Paddlepaddle==2.1.0 edition
- OpenCV-python
- Needed requirements → All the requirements of paddle_detection
- Test system → window10
- The provided pkged enviroment(coda enviroment wich have all the needed libs) --> →Google Drive
Download: →Google Drive → need to put them in the folder named best_model(root_dir)
git clone "https://github.com/ReverseSacle/FairMOT_Paddle.git"
- Unzip paddle-enviroment in Anaconda3/envs/ folder
- Use pycharm,choose paddle-env enviroment.Then,create a folder named best_model,unzip the inference-weight in the best_model folder.
- →Paddle_Detection →
git clone "https://github.com/PaddlePaddle/PaddleDetection.git"
- →Prepare datasets
- →Start to train