Enhancing Visual Reliance in Text Generation: A Bayesian Perspective on Mitigating Hallucination in Large Vision-Language Models

TL;DR

This paper introduces a Bayesian approach to reduce hallucinations in large vision-language models by evaluating visual token informativeness, rectifying prior information, and stopping generation early, leading to improved alignment with visual input.

Contribution

The paper systematically analyzes factors causing hallucination in LVLMs and proposes a Bayesian framework with three strategies to enhance visual reliance in text generation.

Findings

Significant reduction in hallucinations across three benchmarks.

Improved alignment between generated text and visual input.

Outperforms previous state-of-the-art methods in mitigating hallucination.

Abstract

Large Vision-Language Models (LVLMs) usually generate texts which satisfy context coherence but don't match the visual input. Such a hallucination issue hinders LVLMs' applicability in the real world. The key to solving hallucination in LVLM is to make the text generation rely more on the visual content. Most previous works choose to enhance/adjust the features/output of a specific modality (i.e., visual or textual) to alleviate hallucinations in LVLM, which do not explicitly or systematically enhance the visual reliance. In this paper, we comprehensively investigate the factors which may degenerate the visual reliance in text generation of LVLM from a Bayesian perspective. Based on our observations, we propose to mitigate hallucination in LVLM from three aspects. Firstly, we observe that not all visual tokens are informative in generating meaningful texts. We propose to evaluate and remove redundant visual tokens to avoid their disturbance. Secondly, LVLM may encode inappropriate prior information, making it lean toward generating unexpected words. We propose a simple yet effective way to rectify the prior from a Bayesian perspective. Thirdly, we observe that starting from certain steps, the posterior of next-token prediction conditioned on visual tokens may collapse to a prior distribution which does not depend on any informative visual tokens at all. Thus, we propose to stop further text generation to avoid hallucination. Extensive experiments on three benchmarks including POPE, CHAIR, and MME demonstrate that our method can consistently mitigate the hallucination issue of LVLM and performs favorably against previous state-of-the-arts.