Driver-AV Training

This project implements a machine learning model for autonomous vehicle that represents the behaviors from a specific dataset. The communication between the Python backend and the Unity frontend is handled via UDP.

Backend for Deployment (Unity UDP Playback)

The playback.ipynb notebook serves as the backend for real-time inference and communication with a Unity client.

How it Works

The notebook loads the pre-trained model and listens for incoming UDP packets from Unity about the current interation between driver and AV in the virtual environment. When a packet is received, the backend processes the data, runs it through the model for inference, and sends the results back to the Unity client over UDP.

Usage

1. Run the Backend: Open playback.ipynb and run all the cells to start the UDP server.
2. Start the Unity Scene: Open your Unity project and run the scene that contains the UDP client script.
3. Real-time Playback: The Unity application will now communicate with the Python backend, sending data and receiving model predictions for real-time playback and visualization.

Configuration

• IP Address: 127.0.0.1 (localhost)
• Port: 5000 by default

Make sure these settings match in both the playback.ipynb notebook and your Unity client script.