HiVG: Hierarchical Multimodal Fine-grained Modulation for Visual Grounding

TL;DR

HiVG introduces a hierarchical multimodal framework that enhances visual grounding by addressing cross-modal inconsistencies and reducing perceptual errors, leveraging pre-training and low-rank adaptation for improved accuracy and efficiency.

Contribution

The paper proposes HiVG, a novel hierarchical multimodal fine-grained modulation framework that effectively bridges the gap between pre-training and grounding tasks using a multi-layer adaptive cross-modal bridge and HiLoRA.

Findings

Outperforms existing methods on five datasets

Demonstrates significant grounding accuracy improvements

Offers energy-efficient model design

Abstract

Visual grounding, which aims to ground a visual region via natural language, is a task that heavily relies on cross-modal alignment. Existing works utilized uni-modal pre-trained models to transfer visual or linguistic knowledge separately while ignoring the multimodal corresponding information. Motivated by recent advancements in contrastive language-image pre-training and low-rank adaptation (LoRA) methods, we aim to solve the grounding task based on multimodal pre-training. However, there exists significant task gaps between pre-training and grounding. Therefore, to address these gaps, we propose a concise and efficient hierarchical multimodal fine-grained modulation framework, namely HiVG. Specifically, HiVG consists of a multi-layer adaptive cross-modal bridge and a hierarchical multimodal low-rank adaptation (HiLoRA) paradigm. The cross-modal bridge can address the inconsistency between visual features and those required for grounding, and establish a connection between multi-level visual and text features. HiLoRA prevents the accumulation of perceptual errors by adapting the cross-modal features from shallow to deep layers in a hierarchical manner. Experimental results on five datasets demonstrate the effectiveness of our approach and showcase the significant grounding capabilities as well as promising energy efficiency advantages. The project page: https://github.com/linhuixiao/HiVG.