1 paper · 1 filter
Xinmiao Hu, Chun Wang, Ruihe An +4
Multimodal Large Language Models (MLLMs) have demonstrated strong performance in visual understanding tasks, yet they often suffer from object hallucinations--generating descriptio…