Fall Detection

Striking the Balance: Human Pose Estimation based Optimal Fall Recognition

Dependencies

The dependencies are listed in requirements.txt. You can install them with the following command:

pip install -r requirements.txt

Dataset Preparation

• Create a folder called dataset in the root folder of the project. The folder structure should be as follows:

  Fall Detection
  ├── data
  │   ├── adl-01-cam0-urfdd-1-no_keypoints.npy
  │   ├── ...
  ├── remaining files

Training

• The training code uses the PyTorch framework.
• To start training, run the following command:

  python main.py --dataset ./data

• Other arguments can be found in the main.py file.
• The trained models are stored in the ./output folder.
• The hyperparameters of the model can be changed in the ./config/model.json file.