Understanding Retrieval-Augmented Task Adaptation for Vision-Language Models

TL;DR

This paper systematically studies how retrieval-augmented adaptation influences vision-language models, revealing key components and theoretical insights that improve understanding and performance, especially in low-data and unseen category scenarios.

Contribution

It provides a comprehensive analysis of retrieval components in vision-language model adaptation, combining empirical experiments with theoretical foundations to enhance understanding.

Findings

Logit ensemble is critical for effective adaptation.

Cross-modal retrieval plays a vital role in adaptation.

Theoretical support explains empirical observations.

Abstract

Pre-trained contrastive vision-language models have demonstrated remarkable performance across a wide range of tasks. However, they often struggle on fine-trained datasets with categories not adequately represented during pre-training, which makes adaptation necessary. Recent works have shown promising results by utilizing samples from web-scale databases for retrieval-augmented adaptation, especially in low-data regimes. Despite the empirical success, understanding how retrieval impacts the adaptation of vision-language models remains an open research question. In this work, we adopt a reflective perspective by presenting a systematic study to understand the roles of key components in retrieval-augmented adaptation. We unveil new insights on uni-modal and cross-modal retrieval and highlight the critical role of logit ensemble for effective adaptation. We further present theoretical underpinnings that directly support our empirical observations.