DensePose Estimation on COCO (val)
67.77GPS APHCMoCo
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| HCMoCoPre-train Datasets=NTURGBD+COCO, Initialization Strategy=ImageNet pre-trained, Training Data Scale=Full Data2022.03 | 67.77 | 68.29 | 68.63 | |
| HCMoCoPre-train Datasets=NTURGBD+MPII, Initialization Strategy=ImageNet pre-trained, Training Data Scale=Full Data2022.03 | 67.33 | 68.12 | 68.72 | |
| CMCPre-train Datasets=NTURGBD+COCO, Initialization Strategy=ImageNet pre-trained, Training Data Scale=Full Data2022.03 | 67.22 | 67.77 | 68.46 | |
| MMVPre-train Datasets=NTURGBD+MPII, Initialization Strategy=ImageNet pre-trained, Training Data Scale=Full Data2022.03 | 66.67 | 67.51 | 68.29 | |
| IN Pre-trainInitialization Strategy=ImageNet pre-trained, Training Data Scale=Full Data2022.03 | 66.48 | 67.42 | 68.63 | |
| CMCPre-train Datasets=NTURGBD+MPII, Initialization Strategy=ImageNet pre-trained, Training Data Scale=Full Data2022.03 | 66.16 | 67.3 | 68.06 | |
| HCMoCoPre-train Datasets=NTURGBD+MPII, Initialization Strategy=Random initialization, Training Data Scale=Full Data2022.03 | 65.89 | 66.92 | 67.66 | |
| CMCPre-train Datasets=NTURGBD+MPII, Initialization Strategy=Random initialization, Training Data Scale=Full Data2022.03 | 64.97 | 65.66 | 66.96 | |
| MMVPre-train Datasets=NTURGBD+MPII, Initialization Strategy=Random initialization, Training Data Scale=Full Data2022.03 | 64.23 | 65.47 | 67.03 | |
| From ScratchInitialization Strategy=Random initialization, Training Data Scale=Full Data2022.03 | 62.03 | 63.61 | 65.88 | |
| HCMoCoPre-train Datasets=NTURGBD+COCO, Initialization Strategy=ImageNet pre-trained, Training Data Scale=10% Data2022.03 | 54.01 | 56.64 | 59.93 | |
| CMCPre-train Datasets=NTURGBD+COCO, Initialization Strategy=ImageNet pre-trained, Training Data Scale=10% Data2022.03 | 53.53 | 56.18 | 59.37 | |
| HCMoCoPre-train Datasets=NTURGBD+MPII, Initialization Strategy=ImageNet pre-trained, Training Data Scale=10% Data2022.03 | 51.5 | 54.47 | 58.66 | |
| MMVPre-train Datasets=NTURGBD+MPII, Initialization Strategy=ImageNet pre-trained, Training Data Scale=10% Data2022.03 | 50.28 | 53.54 | 58.32 | |
| CMCPre-train Datasets=NTURGBD+MPII, Initialization Strategy=ImageNet pre-trained, Training Data Scale=10% Data2022.03 | 48.82 | 52.57 | 57.94 | |
| HCMoCoPre-train Datasets=NTURGBD+MPII, Initialization Strategy=Random initialization, Training Data Scale=10% Data2022.03 | 48.47 | 51.65 | 56.15 | |
| IN Pre-trainInitialization Strategy=ImageNet pre-trained, Training Data Scale=10% Data2022.03 | 44.34 | 49.11 | 56.11 | |
| CMCPre-train Datasets=NTURGBD+MPII, Initialization Strategy=Random initialization, Training Data Scale=10% Data2022.03 | 43.84 | 47.94 | 54 | |
| MMVPre-train Datasets=NTURGBD+MPII, Initialization Strategy=Random initialization, Training Data Scale=10% Data2022.03 | 41.4 | 45.99 | 52.52 | |
| From ScratchInitialization Strategy=Random initialization, Training Data Scale=10% Data2022.03 | 35.75 | 41.62 | 49.92 |