🚗 Autonomous Navigation in CARLA Using Reinforcement Learning

👨‍💻 Project by:

• Srinivas Peri
• Srimathnath Thejasvi Vondivillu

---

📝 Overview

This project explores the design and training of a Reinforcement Learning (RL) agent for autonomous navigation in the CARLA simulator. Using a Deep Q-Network (DQN), the agent learns to:

• Avoid collisions just using camera feed.
• Adhere to traffic rules and maintain a constant speed of 50 mi/h.

---

🌟 Features

• Simulator: CARLA (v0.9.15) for realistic urban driving scenarios.
• Algorithm: Deep Q-Network (DQN) with Xception-based architecture.
• Dynamic Rewards: Encourages safe, efficient, and rule-compliant navigation.

---

🔗 Repository Includes:

• Code for training and evaluating the RL agent.
• Preprocessed data and logs.
• Detailed project report: Report Link.