Multimodal Large Language Models (MLLMs) have achieved SOTA performance in various visual language tasks by fusing the visual representations with LLMs leveraging some visual adapters. In this paper, we first establish that adapters using query-based Transformers such as Q-former is a simplified Multi-instance Learning method without considering instance heterogeneity/correlation. We then propose a general component termed Multi-instance Visual Prompt Generator (MIVPG) to incorporate enriched visual representations into LLMs by taking advantage of instance correlation between images or patches for the same sample. Quantatitive evaluation on three public vision-language (VL) datasets from different scenarios shows that the proposed MIVPG improves Q-former in main VL tasks.
@article{arxiv.2406.02987,
title = {Enhancing Multimodal Large Language Models with Multi-instance Visual Prompt Generator for Visual Representation Enrichment},
author = {Wenliang Zhong and Wenyi Wu and Qi Li and Rob Barton and Boxin Du and Shioulin Sam and Karim Bouyarmane and Ismail Tutar and Junzhou Huang},
journal= {arXiv preprint arXiv:2406.02987},
year = {2024}
}