V-DPO: Mitigating Hallucination in Large Vision Language Models via Vision-Guided Direct Preference Optimization
Vision-guided Direct Preference Optimization (V-DPO) addresses hallucination issues in large vision-language models (LVLMs) where text responses can stray from visual input due to an over-reliance on language.
#technology #artificilaintelligence #ai #vdpo #lvlm #hallucination