Hierarchical Alignment-enhanced Adaptive Grounding Network for Generalized Referring Expression Comprehension

TL;DR

This paper introduces HieA2G, a novel hierarchical alignment and adaptive grounding network that significantly improves generalized referring expression comprehension by handling complex, multi-target, and no-target cases with state-of-the-art performance.

Contribution

The paper proposes a Hierarchical Multi-modal Semantic Alignment module and an Adaptive Grounding Counter, enabling flexible, comprehensive understanding and dynamic target counting in GREC.

Findings

Achieves state-of-the-art results on GREC and related tasks.

Enhances multi-modal understanding through hierarchical alignment.

Effectively handles varying target numbers with adaptive grounding.

Abstract

In this work, we address the challenging task of Generalized Referring Expression Comprehension (GREC). Compared to the classic Referring Expression Comprehension (REC) that focuses on single-target expressions, GREC extends the scope to a more practical setting by further encompassing no-target and multi-target expressions. Existing REC methods face challenges in handling the complex cases encountered in GREC, primarily due to their fixed output and limitations in multi-modal representations. To address these issues, we propose a Hierarchical Alignment-enhanced Adaptive Grounding Network (HieA2G) for GREC, which can flexibly deal with various types of referring expressions. First, a Hierarchical Multi-modal Semantic Alignment (HMSA) module is proposed to incorporate three levels of alignments, including word-object, phrase-object, and text-image alignment. It enables hierarchical cross-modal interactions across multiple levels to achieve comprehensive and robust multi-modal understanding, greatly enhancing grounding ability for complex cases. Then, to address the varying number of target objects in GREC, we introduce an Adaptive Grounding Counter (AGC) to dynamically determine the number of output targets. Additionally, an auxiliary contrastive loss is employed in AGC to enhance object-counting ability by pulling in multi-modal features with the same counting and pushing away those with different counting. Extensive experimental results show that HieA2G achieves new state-of-the-art performance on the challenging GREC task and also the other 4 tasks, including REC, Phrase Grounding, Referring Expression Segmentation (RES), and Generalized Referring Expression Segmentation (GRES), demonstrating the remarkable superiority and generalizability of the proposed HieA2G.