Mitigating Object Hallucinations in Large Vision-Language Models through Visual Contrastive Decoding | Read Paper on Bytez