Scene Parsing on NYUDv2 (test)
60.8mIoUOmniVec
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| OmniVecInput Data=RGB-D2024.04 | 60.8 | — | — | 1 | |
| DFormer-LInput Data=RGB-D2024.04 | 57.2 | — | — | 5 | |
| CMXInput Data=RGB-HHA2024.04 | 56.9 | — | — | 6 | |
| CMNeXtInput Data=RGB-HHA2024.04 | 56.9 | — | — | 6 | |
| OmnivoreInput Data=RGB-D2024.04 | 56.8 | — | — | 8 | |
| HAPNetInput Data=RGB-HHA2024.04 | 55 | 79.2 | 68.8 | 15 | |
| TokenFusionInput Data=RGB-D2024.04 | 54.2 | 79 | 66.9 | 16 | |
| AsymFormerInput Data=RGB-D2024.04 | 54.1 | 78.5 | — | 17 | |
| SA-GateInput Data=RGB-HHA2024.04 | 52.4 | — | — | 31 | |
| ECGFNetInput Data=RGB-D2024.04 | 51.5 | — | 65.2 | 37 | |
| Malleable 2.5DBackbone=ResNet-101 x1, Multi-scale and flipping inference=true2020.07 | 50.9 | 76.9 | — | — | |
| SANet2020.11 | 50.7 | 75.9 | — | — | |
| IdempotentBackbone=ResNet-101 x22020.07 | 50.6 | 76.3 | — | — | |
| Pad-Net2020.11 | 50.2 | 75.2 | — | — | |
| PAD-NetTraining strategy=Two-stage2021.01 | 50.2 | 75.2 | 0.623 | — | |
| RDF-152Backbone=ResNet-152 x22020.07 | 50.1 | 76 | — | — | |
| Malleable 2.5DBackbone=ResNet-50 x1, Multi-scale and flipping inference=true2020.07 | 49.7 | 76.3 | — | — | |
| 2.5D+HHABackbone=ResNet-101 x22020.07 | 49.1 | 75.9 | — | — | |
| RDF-101Backbone=ResNet-101 x22020.07 | 49.1 | 75.6 | — | — | |
| RTFNetInput Data=RGB-D2024.04 | 49.1 | — | 64.8 | 59 | |
| MSCI2020.11 | 49 | — | — | — | |
| 2.5DBackbone=ResNet-101 x12020.07 | 48.4 | 75.3 | — | — | |
| CFN(RefineNet-152)Backbone=ResNet-152 x22020.07 | 47.7 | — | — | — | |
| RefineNet2020.11 | 47.6 | 74.4 | — | — | |
| RefineNetTraining strategy=One-stage2021.01 | 46.5 | 73.6 | 0.589 | — | |
| ESOSD-NetTraining strategy=One-stage2021.01 | 45 | 73.3 | 0.647 | — | |
| Depth-aware+HHABackbone=VGG-16 x22020.07 | 43.9 | — | — | — | |
| SEGCloud2020.11 | 43.5 | — | — | — | |
| Lin et al.2020.11 | 40.6 | 70 | — | — | |
| ContextTraining strategy=One-stage2021.01 | 40.6 | 70 | 0.536 | — | |
| Arsalan et al.Training strategy=Two-stage2021.01 | 39.2 | 68.6 | 0.523 | — | |
| Eigen and FergusTraining strategy=One-stage2021.01 | 34.1 | 65.6 | 0.451 | — | |
| FCN+HHABackbone=VGG-16 x22020.07 | 34 | 65.4 | — | — | |
| FCNTraining strategy=One-stage2021.01 | 29.2 | 60 | 0.422 | — | |
| Gupta et al.Training strategy=Two-stage2021.01 | 28.6 | 60.3 | — | — | |
| Deng et al.Training strategy=One-stage2021.01 | — | 63.8 | 0.315 | — |