Скачать или смотреть DQN + LLM Drone Navigation (Fundamentals of AI Agents)

Demonstration of an autonomous drone navigation system that combines Deep Q-Network (DQN) reinforcement learning with a Large Language Model (LLM)-driven navigator in a procedurally generated forest environment. This project was developed as part of the “Fundamentals of AI Agents” course under Prof. Ahmed Esmaeili, in collaboration with Mohammed Farhan and Mohammed Rizwan Raisulhaq.

Key features:
• DQN-based policy training and evaluation in PyBullet
• LLM-powered waypoint generation for adaptive navigation
• Interactive GUI for visualizing agent performance

To reproduce or explore the code, visit the repository:
https://github.com/qasim418/intellige...

Setup:
• Install dependencies: pip install -r requirements.txt
• Train: python train_dqn_random_nav.py
• Evaluate: python evaluate_trained_policy_gui.py
• Set OPENROUTER_API_KEY in a .env file for LLM integration