HierSum: A Global and Local Attention Mechanism for Video Summarization
This work addresses the problem of efficiently summarizing instructional videos for users needing quick overviews, though it is incremental as it builds on existing methods with a novel hybrid approach.
The authors tackled video summarization for instructional videos by proposing HierSum, a hierarchical method that combines local subtitle cues with global video-level instructions and uses 'most replayed' statistics as supervision, resulting in consistent outperformance of existing methods on benchmark datasets like TVSum and BLiSS in metrics such as F1-score and rank correlation.
Video summarization creates an abridged version (i.e., a summary) that provides a quick overview of the video while retaining pertinent information. In this work, we focus on summarizing instructional videos and propose a method for breaking down a video into meaningful segments, each corresponding to essential steps in the video. We propose \textbf{HierSum}, a hierarchical approach that integrates fine-grained local cues from subtitles with global contextual information provided by video-level instructions. Our approach utilizes the ``most replayed" statistic as a supervisory signal to identify critical segments, thereby improving the effectiveness of the summary. We evaluate on benchmark datasets such as TVSum, BLiSS, Mr.HiSum, and the WikiHow test set, and show that HierSum consistently outperforms existing methods in key metrics such as F1-score and rank correlation. We also curate a new multi-modal dataset using WikiHow and EHow videos and associated articles containing step-by-step instructions. Through extensive ablation studies, we demonstrate that training on this dataset significantly enhances summarization on the target datasets.