Panoptic Segmentation on NYU v2 (test)
51.15PQPanopticNDT (EMSANet-R34-NBt1D)
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| PanopticNDT (EMSANet-R34-NBt1D)Protocol=Fine-tuning, LR=0.004, Selection=Best in run2023.09 | 51.15 | 57.9 | 60.04 | 84.99 | 13.86 | |
| PanopticNDT (EMSANet-R34-NBt1D)Protocol=Application network, LR=0.00052023.09 | 48.47 | 55.42 | 56.77 | 84.27 | 14.58 | |
| EMSANetBackbone=2xR34-Nbt1d, Modality=RGB-D, Pre-training=Hypersim2022.07 | 47.38 | — | — | — | — | |
| EMSANetBackbone=2x ResNet101, Task(s)=Sem + Sce + Ins + Or2023.06 | 45.12 | 50.67 | 54.02 | 82.49 | 15.33 | |
| EMSANetBackbone=2xR101-Nbt1d, Modality=RGB-D2022.07 | 45.12 | — | — | — | — | |
| EMSANetBackbone=2x ResNet34-NBt1D, Task(s)=Sem + Sce + Ins + Or2023.06 | 43.59 | 50.61 | 52.23 | 82.48 | 16.39 | |
| EMSANetBackbone=2xR34-Nbt1d, Modality=RGB-D2022.07 | 43.56 | — | — | — | — | |
| EMSAFormerBackbone=SwinV2-T-128-Multi-Aug, Task(s)=Sem(SegFormer) + Sce + Ins + Or, Semantic Decoder=SegFormer2023.06 | 43.41 | 51.34 | 52.53 | 81.75 | 18.94 | |
| EMSAFormerBackbone=SwinV2-T-128-Multi-Aug, Task(s)=Sem + Sce + Ins + Or, Semantic Decoder=EMSANet2023.06 | 43.28 | 51.76 | 52.48 | 81.43 | 18.26 | |
| EMSANetBackbone=R34-Nbt1d, Modality=RGB2022.07 | 37.69 | — | — | — | — | |
| Panoptic DeepLabBackbone=R101, Modality=RGB, Enriched NYUv2=true2022.07 | 35.32 | — | — | — | — | |
| Panoptic DeepLabBackbone=R50, Modality=RGB, Enriched NYUv2=true2022.07 | 30.99 | — | — | — | — |