VLALight: A Vision-Language-Action Model for Traffic Signal Control
VLALight is presented, the first vision-language-action (VLA) model for end-to-end traffic signal control from multi-view roadside videos that consistently outperforms transportation-based, RL-based, and LLM/VLM-based baselines.
Pan Zhang, Si-Qi Lai, Ke-Mu Dong et al.
· 0 citations