This project implements a visual perception system for autonomous driving simulation using deep learning techniques. It combines lane detection (using UFLD) and traffic signal detection (using YOLOv8n) , suitable for research, prototyping, and experimentation.