Mask What Matters: Mitigating Object Hallucinations in Multimodal Large Language Models with Object-Aligned Visual Contrastive Decoding.

Boqi Chen, Xudong Liu, Jianing Qiu

Browse the full EACL paper archive.