Traffic Sign Detection and Recognition using G-CNNs for Autonomous Vehicles

Traffic sign detection is a crucial task for autonomous vehicles, enabling safe navigation through complex environments. Traditional Convolutional Neural Networks (CNNs) are highly effective in detecting traffic signs, but their computational complexity and large memory requirements make them challenging to deploy in real-time systems, especially on resource-limited devices. Binary Neural Networks (BNNs) provide a more efficient alternative by simplifying computations and reducing memory usage, enabling faster processing, but they often sacrifice the detailed features that CNNs capture. Group Convolutional Neural Networks (GCNNs), on the other hand, excel at understanding spatial relationships, such as how traffic signs relate to their surrounding environment and road networks. By combining the computational efficiency of BNNs with the spatial awareness of G-CNNs, this approach aims to deliver a solution that balances speed, efficiency, and the ability to detect traffic signs with contextual understanding.