Semantic Segmentation on LIP (test)
56.48mIoUHRNetV2 + OCR
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| HRNetV2 + OCRbackbone=HRNetV2-W48, extra=N2019.08 | 56.48 | 88.24 | 67.84 | |
| HRNetV2backbone=HRNetV2-W48, extra=N2019.08 | 55.9 | 88.21 | 67.43 | |
| CE2Pbackbone=D-ResNet-101, extra=Edge2019.08 | 53.1 | 87.37 | 63.2 | |
| JPPNetbackbone=D-ResNet-101, extra=Pose2019.08 | 51.37 | 86.39 | 62.32 | |
| MULAbackbone=Hourglass, extra=Pose2019.08 | 49.3 | 88.5 | 60.5 | |
| SS-NANbackbone=ResNet-101, extra=Pose2019.08 | 47.92 | 87.59 | 56.03 | |
| MMANbackbone=D-ResNet-1012019.08 | 46.81 | — | — | |
| DeepLabV3+backbone=D-ResNet-1012019.08 | 44.8 | 84.09 | 55.62 | |
| Attention+SSLbackbone=VGG16, extra=Pose2019.08 | 44.73 | 84.36 | 54.94 |