Jun 21, 2022 · Abstract:Video highlight detection is a crucial yet challenging problem that aims to identify the interesting moments in untrimmed videos.

A supervised hard-pairs guided contrastive learning scheme is deployed to reflect structural representation of video sequence. Besides, a hard-pairs mining ...

The key to this task lies in effective video representations that jointly pursue two goals, i.e., 1) cross-modal representation learning and 2) fine-grained ...

Video highlight detection is a crucial yet challenging problem that aims to identify the interesting moments in untrimmed videos.

[2022] Probing Visual-Audio Representation for Video Highlight Detection via Hard-Pairs Guided Contrastive Learning Authors: Shuaicheng Li, Feng Zhang ...

Probing Visual-Audio Representation for Video Highlight Detection via Hard-Pairs Guided Contrastive Learning ... Our proposed method mainly leverages the intra- ...

Probing visual-audio representation for video highlight detection via hard-pairs guided contrastive learning. S Li, F Zhang, K Yang, L Liu, S Liu, J Hou, S ...

Probing Visual-Audio Representation for Video Highlight Detection via Hard-Pairs Guided Contrastive Learning · no code implementations • 21 Jun 2022 ...

This work proposes a novel triplet deep ranking approach to video highlight detection using Web images as a weak supervision, which is fully category ...

Probing Visual-Audio Representation for Video Highlight Detection via Hard-Pairs Guided Contrastive Learning. Preprint. Jun 2022. Shuaicheng Li ...