SHE-Net: Syntax-Hierarchy-Enhanced Text-Video Retrieval

TL;DR

Addresses the modality gap in text-video retrieval by leveraging textual syntax to guide frame and region selection in videos and by guiding cross-modal similarity computation. The proposed SHE-Net builds a text syntax hierarchy that reveals grammatical structure and uses it to induce a complementary video syntax hierarchy, enabling syntax-guided multimodal fusion and alignment. Through experiments and ablations on four public datasets (MSR-VTT, MSVD, DiDeMo, ActivityNet), the method demonstrates notable improvements over baselines. This approach highlights the value of linguistic structure in multi-modal retrieval, offering finer-grained cross-modal understanding and more accurate video retrieval from text queries.

Abstract

The user base of short video apps has experienced unprecedented growth in recent years, resulting in a significant demand for video content analysis. In particular, text-video retrieval, which aims to find the top matching videos given text descriptions from a vast video corpus, is an essential function, the primary challenge of which is to bridge the modality gap. Nevertheless, most existing approaches treat texts merely as discrete tokens and neglect their syntax structures. Moreover, the abundant spatial and temporal clues in videos are often underutilized due to the lack of interaction with text. To address these issues, we argue that using texts as guidance to focus on relevant temporal frames and spatial regions within videos is beneficial. In this paper, we propose a novel Syntax-Hierarchy-Enhanced text-video retrieval method (SHE-Net) that exploits the inherent semantic and syntax hierarchy of texts to bridge the modality gap from two perspectives. First, to facilitate a more fine-grained integration of visual content, we employ the text syntax hierarchy, which reveals the grammatical structure of text descriptions, to guide the visual representations. Second, to further enhance the multi-modal interaction and alignment, we also utilize the syntax hierarchy to guide the similarity calculation. We evaluated our method on four public text-video retrieval datasets of MSR-VTT, MSVD, DiDeMo, and ActivityNet. The experimental results and ablation studies confirm the advantages of our proposed method.