1 paper · 1 filter
Yike Wu, Yiwei Wang, Yujun Cai
While Large Vision-Language Models (LVLMs) achieve strong performance in multimodal tasks, hallucinations continue to hinder their reliability. Among the three categories of halluc…