Siamese Learning with Joint Alignment and Regression for Weakly-Supervised Video Paragraph Grounding

arXiv:2403.11463v1 Announce Type: new
Abstract: Video Paragraph Grounding (VPG) is an emerging task in video-language understanding, which aims at localizing multiple sentences with semantic relations and temporal order from an untrimmed video. However, existing VPG approaches are heavily reliant on a considerable number of temporal labels that are laborious and time-consuming to acquire. In this work, we introduce and explore Weakly-Supervised Video Paragraph Grounding (WSVPG) to eliminate the need of temporal annotations. Different from previous weakly-supervised grounding frameworks based on …

abstract alignment arxiv cs.cv however labels language language understanding multiple regression relations semantic temporal type understanding video weakly-supervised