
Progressive Contrastive Learning mit Multi-Prototypen verbessert die Leistung der unsupervised Visible-Infrared Person Re-identification.


coremsg

progressive-contrastive-learning-with-multi-prototype-for-unsupervised-visible-infrared-person-re-identification


Progressive Contrastive Learning with Multi-Prototype for Unsupervised Visible-Infrared Person Re-identification


title_rewrite


Progressive Contrastive Learning with Multi-Prototype (PCLMP) method enhances Unsupervised Visible-Infrared Person Re-identification by incorporating hard and dynamic prototypes.
