Dual-Branch Occlusion-Aware Semantic Part-Features Extraction Network for Occluded Person Re-Identification

Occlusion remains a major challenge in person re-identification, as it often leads to incomplete or misleading visual cues. To address this issue, we propose a dual-branch occlusion-aware network (DOAN), which explicitly and implicitly enhances the model’s capability to perceive and handle occlusion...

Ausführliche Beschreibung

Bibliografische Detailangaben

Hauptverfasser:	Bo Sun, Yulong Zhang, Jianan Wang, Chunmao Jiang
Format:	Artigo
Sprache:	Inglês
Veröffentlicht:	MDPI AG 2025-07-01
Schriftenreihe:	Mathematics
Schlagworte:	occluded person re-identification part-aware feature learning occlusion data augmentation attention mechanism feature recovery
Online-Zugang:	https://www.mdpi.com/2227-7390/13/15/2432
Tags:	Tag hinzufügen Keine Tags, Fügen Sie das erste Tag hinzu!