1 paper · 1 filter
Le Yang, Ziwei Zheng, Boxu Chen +3
Recent studies have shown that large vision-language models (LVLMs) often suffer from the issue of object hallucinations (OH). To mitigate this issue, we introduce an efficient met…