Dual-level aggregation network for video-based visible-infrared group re-identification

Open Webpage

Jianghao Xiong, Xiaohua Xie, Jian-Huang Lai

Published: 2026, Last Modified: 19 Oct 2025Pattern Recognit. 2026EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•We propose a dual-level aggregation module for robust features in video group re-id.•We propose an asymptotic modality learning strategy to reduce modality gap.•We release a large-scale video-based visible-infrared group re-identification dataset.•Our method achieves state-of-the-art results in video visible-infrared group re-id.