Role of Audio in Audio-Visual Video Summarization

TL;DR

This paper introduces a novel audio-visual video summarization framework that leverages multiple fusion methods and explainability techniques, demonstrating improved performance on benchmark datasets.

Contribution

It proposes a new audio-visual fusion approach using GRU and attention networks, along with an explainability method based on audio-visual CCA to enhance understanding of audio's role.

Findings

Improved F1 and Kendall-tau scores on TVSum dataset.

Enhanced performance for positively correlated audio-visual content.

Effective use of CCA for explainability in video summarization.

Abstract

Video summarization attracts attention for efficient video representation, retrieval, and browsing to ease volume and traffic surge problems. Although video summarization mostly uses the visual channel for compaction, the benefits of audio-visual modeling appeared in recent literature. The information coming from the audio channel can be a result of audio-visual correlation in the video content. In this study, we propose a new audio-visual video summarization framework integrating four ways of audio-visual information fusion with GRU-based and attention-based networks. Furthermore, we investigate a new explainability methodology using audio-visual canonical correlation analysis (CCA) to better understand and explain the role of audio in the video summarization task. Experimental evaluations on the TVSum dataset attain F1 score and Kendall-tau score improvements for the audio-visual video summarization. Furthermore, splitting video content on TVSum and COGNIMUSE datasets based on audio-visual CCA as positively and negatively correlated videos yields a strong performance improvement over the positively correlated videos for audio-only and audio-visual video summarization.