A Lightweight CNN-Based Model for Real-Time Traffic Sign Recognition in Autonomous Vehicles

The rapid advancement of autonomous vehicles demands fast and accurate perception systems capable of recognizing traffic signs in real-time.Traditional convolutional neural networks (CNNs) provide excellent accuracy but are computationally expensive for embedded devices.This research presents TrafficSignNet -a novel lightweight CNN model designed for efficient and real-time traffic sign recognition.The proposed architecture is optimized for edge hardware such as NVIDIA Jetson devices.Experimental evaluation demonstrates that TrafficSignNet achieves a mean accuracy of 96.4% with a model size under 8 MB and inference speed of 85 FPS, making it ideal for deployment in autonomous driving environments.