BMIP: Bi-directional Modality Interaction Prompt Learning for VLM

TL;DR

This paper introduces BMIP, a novel bi-directional modality interaction prompt learning method for vision-language models that enhances inter-modal alignment and generalization across diverse tasks.

Contribution

The paper proposes BMIP, a new prompt learning approach that dynamically weights bi-modal information for improved inter-modal interaction and generalization in VLMs.

Findings

BMIP outperforms state-of-the-art methods across multiple evaluation paradigms.

BMIP enhances inter-modal consistency and trainability.

Flexible integration with other prompt methods improves performance.

Abstract

Vision-language models (VLMs) have exhibited remarkable generalization capabilities, and prompt learning for VLMs has attracted great attention for the ability to adapt pre-trained VLMs to specific downstream tasks. However, existing studies mainly focus on single-modal prompts or uni-directional modality interaction, overlooking the powerful alignment effects resulting from the interaction between the vision and language modalities. To this end, we propose a novel prompt learning method called $\underline{\textbf{B}}i-directional \underline{\textbf{M}}odality \underline{\textbf{I}}nteraction \underline{\textbf{P}}rompt (BMIP)$, which dynamically weights bi-modal information through learning the information of the attention layer, enhancing trainability and inter-modal consistency compared to simple information aggregation methods. To evaluate the effectiveness of prompt learning methods, we propose a more realistic evaluation paradigm called open-world generalization complementing the widely adopted cross-dataset transfer and domain generalization tasks. Comprehensive experiments on various datasets reveal that BMIP not only outperforms current state-of-the-art methods across all three evaluation paradigms but is also flexible enough to be combined with other prompt-based methods for consistent performance enhancement.