AI RESEARCH
Pedestrian Crossing Intent Prediction via Psychological Features and Transformer Fusion
arXiv CS.CV
•
ArXi:2603.19533v1 Announce Type: new Pedestrian intention prediction needs to be accurate for autonomous vehicles to navigate safely in urban environments. We present a lightweight, socially informed architecture for pedestrian intention prediction. It fuses four behavioral streams (attention, position, situation, and interaction) using highway encoders, a compact 4-token Transformer, and global self-attention pooling.