Adaptive Hierarchical Graph Reasoning with Semantic Coherence for Video-and-Language Inference

TL;DR

This paper introduces an adaptive hierarchical graph network with semantic coherence learning to improve video-and-language inference by jointly reasoning over video, subtitles, and complex social interactions.

Contribution

It proposes a novel adaptive hierarchical graph model that adjusts based on semantic structures and incorporates semantic coherence learning for better reasoning.

Findings

Significantly outperforms baseline methods.

Effectively models long-range relationships and social interactions.

Improves alignment between vision and language.

Abstract

Video-and-Language Inference is a recently proposed task for joint video-and-language understanding. This new task requires a model to draw inference on whether a natural language statement entails or contradicts a given video clip. In this paper, we study how to address three critical challenges for this task: judging the global correctness of the statement involved multiple semantic meanings, joint reasoning over video and subtitles, and modeling long-range relationships and complex social interactions. First, we propose an adaptive hierarchical graph network that achieves in-depth understanding of the video over complex interactions. Specifically, it performs joint reasoning over video and subtitles in three hierarchies, where the graph structure is adaptively adjusted according to the semantic structures of the statement. Secondly, we introduce semantic coherence learning to explicitly encourage the semantic coherence of the adaptive hierarchical graph network from three hierarchies. The semantic coherence learning can further improve the alignment between vision and linguistics, and the coherence across a sequence of video segments. Experimental results show that our method significantly outperforms the baseline by a large margin.