CLIP-Guided Adaptable Self-Supervised Learning for Human-Centric Visual Tasks

Benchmark Model Rank Results
person-re-identification-on-msmt17CLASP (TransReID, Swin-B)–mAP: 78.8Rank-1: 91.8
person-re-identification-on-msmt17CLASP (TransReID, Swin-S)–mAP: 77.2Rank-1: 91.2
person-re-identification-on-msmt17CLASP (TransReID, Swin-T)–mAP: 68.1Rank-1: 86.8