2605.28604v1 May 27, 2026 cs.CV

Mining Multi-Modality Spatio-Temporal Cues for Video Important Person Identification

Wenke Huang
Wenke Huang
Citations: 1,958
h-index: 19
Bin Yang
Bin Yang
Citations: 27
h-index: 3
Xiao Wang
Xiao Wang
Citations: 404
h-index: 10
Minglei Yang
Minglei Yang
Citations: 14
h-index: 2
Zheng Wang
Zheng Wang
Citations: 29
h-index: 3
Xin Xu
Xin Xu
Citations: 13
h-index: 3
Mang Ye
Mang Ye
Citations: 20
h-index: 2

Identifying key individuals in video scenes is essential for applications such as automated video editing and intelligent surveillance. Current methods primarily focus on static images and immediate visual cues, overlooking the rich spatio-temporal information in videos. This leads to the phenomenon of Temporal Importance Shift (TIS), wherein individuals deemed significant in early frames may be demoted as the entire temporal context is considered. To address this, we introduce the Video Important Person (VIP) identification task, aimed at automatically identifying the most influential individuals in videos while providing textual rationales. We present Temporal-VIP, a large-scale rationale-annotated dataset consisting of 9,249 video segments across 11 categories with aligned importance rationales. To mitigate TIS, we develop the VIP-Net framework, which includes a Social Cue Encoder (SCE) for extracting multi-modal spatio-temporal cues, a Temporal Importance Rectifier (TIR) for hierarchical cue fusion and cross-modal alignment, and VIP Inference for ranking individuals. Experimental results show that VIP-Net achieves 67.3% accuracy, significantly outperforming state-of-the-art models (37.5%-53.9%) and yielding a mean rationale similarity of 0.63 to ground truth through feature-guided LLM refinement. The dataset and code are available at https://huggingface.co/datasets/yml2002/Temporal-VIP.

0 Citations
0 Influential
29.5 Altmetric
147.5 Score
Original PDF
0

No Analysis Report Yet

This paper hasn't been analyzed by Gemini yet.

Log in to request an AI analysis.

댓글

댓글을 작성하려면 로그인하세요.

아직 댓글이 없습니다. 첫 번째 댓글을 남겨보세요!