AIFA is a real-time, AI-powered system that assists user in correcting exercise postures.
We provide a Python package for setting up the system locally.
- Python 3.10
To get the AI running, follow these steps:
- Clone the repo
git clone https://github.com/tokudayo/AIFA.git
- Setup Python environment
pip install -r ai/requirements.txt
For evaluation, use either eval.py or webcam.py. Here is an example of using eval.py to evaluate:
python -m scripts.evalFor more details, please refer to the AI Documentation.
Here is an example of the system in action. The system will track the user's pose and provide feedback should there be any errors.
To deploy the software, you need to have the following installed:
- WSL 2 (required if you are using Windows)
- Docker 20.10.17
To get the software running with Docker, follow these steps:
- Clone the repo
git clone https://github.com/tokudayo/AIFA.git
- Rename
.env.exampleto.envand fill in the required information. - Build the images and run the container.
docker-compose up -d
For more details, please refer to the Software Documentation.
We provided a demo video for the software. Please refer to the Demo Video.
Distributed under the BSD 3-Clause License License. See LICENSE for more information.