1 paper
Baojie Chen, Zijun Jia, Jing Zhong
VLMs have shown promise for autonomous driving, yet still suffer from hallucination, weak spatio-temporal perception, and limited generalization. Recent methods improve reasoning a…