/predictive-state-representations-ad

Primary LanguagePython

Predictive State Representation Learning for Autonomous Vehicles

Overview

This project develops and evaluates predictive models for autonomous vehicle navigation using deep learning. The goal is to learn ego-centric representations of the future environmental states. It uses CommonRoad-Geometric as the autonomous driving software environment.

Usage

  1. Collect dataset:

    python collect_dataset.py

  2. Train representation model:

    python train_model.py

  3. Train downstream RL agent:

    python train_rl_agent.py

Modify config.yaml to adjust simulation parameters and training settings.

Local Configuration

For machine-specific settings, create a config.local.yaml file based on the config.local.template.yaml:

cp config.local.template.yaml config.local.yaml