This project aims to develop a person re-identification system from a top-view camera using depth files from the TVPR2 dataset. The goal is to leverage this data to create a robust model capable of recognizing individuals in different video sequences.
The `Process_Database.ipynb` notebook is dedicated to cleaning the data from the TVPR2 dataset. It includes necessary steps for preprocessing images, handling annotations, and ensuring the quality of data used for model training.
In the `Process_Sequence_Normalization.ipynb` notebook, you'll find processes for normalizing sequences. This includes normalizing depth values and other transformations needed to prepare data before introducing it to the model.
The `Graph_Creation_Process.ipynb` notebook is specially designed for creating graphs from each image. These graphs are used to train a Graph Convolutional Network (GCN) to enhance re-identification performance.
-
Clone the repository:
git clone https://github.com/your-username/project-name.git
-
Install dependencies:
pip install -r requirements.txt
-
Run the notebooks in the specified order above.
If you want to contribute to the improvement of this project, please follow these steps:
- Fork the project.
- Create a branch for your feature (`git checkout -b feature/add-new-feature`).
- Commit your changes (`git commit -m 'Add a new feature'`).
- Push the branch (`git push origin feature/add-new-feature`).
- Open a Pull Request.
- Simon GARY - Lead Developer
- Ombeline MOINEAU - Lead Developer
This project is licensed under the MIT License.