SURf: Teaching Large Vision-Language Models to Selectively Utilize Retrieved Information

TL;DR

This paper introduces SURf, a self-refinement framework that trains large vision-language models to selectively utilize relevant retrieved information, significantly improving their accuracy and robustness across multiple tasks and datasets.

Contribution

The paper presents a novel self-refinement approach that teaches LVLMs to distinguish and use relevant references, addressing limitations of previous methods.

Findings

Enhanced LVLM performance across three tasks and seven datasets.

Improved robustness against irrelevant or misleading references.

Effective fine-tuning method for selective information utilization.

Abstract

Large Vision-Language Models (LVLMs) have become pivotal at the intersection of computer vision and natural language processing. However, the full potential of LVLMs Retrieval-Augmented Generation (RAG) capabilities remains underutilized. Existing works either focus solely on the text modality or are limited to specific tasks. Moreover, most LVLMs struggle to selectively utilize retrieved information and are sensitive to irrelevant or misleading references. To address these challenges, we propose a self-refinement framework designed to teach LVLMs to Selectively Utilize Retrieved Information (SURf). Specifically, when given questions that are incorrectly answered by the LVLM backbone, we obtain references that help correct the answers (positive references) and those that do not (negative references). We then fine-tune the LVLM backbone using a combination of these positive and negative references. Our experiments across three tasks and seven datasets demonstrate that our framework significantly enhances LVLMs ability to effectively utilize retrieved multimodal references and improves their robustness against irrelevant or misleading information. The source code is available at https://github.com/GasolSun36/SURf.