Semantic Segmentation on SUN-RGBD
54IoURSGMamba-B
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| RSGMamba-BBackbone=SegMAN-B, Params.=48.6M, Input size=530 × 730, Flops=168.4G2026.04 | 54 | — | — | |
| DFormerV2Backbone=DFormerV2-L, Params.=95.5M, Input size=530 × 730, Flops=160.5G2026.04 | 53.3 | — | — | |
| RSGMamba-SBackbone=SegMAN-S, Params.=28.1M, Input size=530 × 730, Flops=75.8G2026.04 | 52.9 | — | — | |
| DiffPixelFormerBackbone=MiT-B3, Params.=85.4M, Input size=530 × 730, Flops=205.7G2026.04 | 52.8 | — | — | |
| GeminiFusionBackbone=MiT-B3, Params.=75.8M, Input size=530 × 7302026.04 | 52.7 | — | — | |
| ADBNetBackbone=ConvNext-base, Params.=299.1M, Input size=530 × 7302026.04 | 52.7 | — | — | |
| PrimKDBackbone=MiT-B4, Params.=139.9M, Input size=530 × 730, Flops=173.8G2026.04 | 52.5 | — | — | |
| CMXBackbone=MiT-B5, Params.=181.1M, Input size=530 × 730, Flops=217.6G2026.04 | 52.4 | — | — | |
| DFNetBackbone=MiT-B4, Params.=108.8M, Input size=530 × 7302026.04 | 52.4 | — | — | |
| SigmaBackbone=VMamba-S, Params.=69.8M, Input size=480 × 640, Flops=139.8G2026.04 | 52.4 | — | — | |
| CMXBackbone=MiT-B4, Params.=139.9M, Input size=530 × 730, Flops=173.8G2026.04 | 52.1 | — | — | |
| CMNextBackbone=MiT-B4, Params.=119.6M, Input size=530 × 730, Flops=170.3G2026.04 | 51.9 | — | — | |
| DFNetBackbone=MiT-B2, Params.=41.2M, Input size=530 × 7302026.04 | 51.6 | — | — | |
| TokenFusionBackbone=MiT-B3, Params.=45.9M, Input size=530 × 730, Flops=122.1G2026.04 | 51 | — | — | |
| Pattern-Affinitive Propagationdata=RGB, Backbone=ResNet-502019.06 | 50.5 | 83.8 | 58.4 | |
| ECMRNBackbone=DFormer, Params.=68.6M, Input size=530 × 730, Flops=178.5G2026.04 | 50.5 | — | — | |
| TRL-ResNet50data=RGB, Backbone=ResNet-502019.06 | 50.3 | 83.6 | 58.9 | |
| SigmaBackbone=VMamba-T, Params.=48.3M, Input size=480 × 640, Flops=90.4G2026.04 | 50 | — | — | |
| DCANetBackbone=VMamba, Params.=123.8M, Input size=530 × 7302026.04 | 49.6 | — | — | |
| ADBNetBackbone=ConvNext-tiny, Params.=45.9M, Input size=530 × 7302026.04 | 49.6 | — | — | |
| CFNdata=RGBD2019.06 | 48.1 | — | — | |
| RDF-152data=RGBD, Backbone=ResNet-1522019.06 | 47.7 | 81.5 | 60.1 | |
| 3D-GNNdata=RGBD2019.06 | 45.9 | — | 57 | |
| RefineNet-101data=RGB, Backbone=ResNet-1012019.06 | 45.7 | 80.4 | 57.8 | |
| Contextdata=RGB2019.06 | 42.3 | 78.4 | 53.4 | |
| DINOv3Backbone=ViT-L/162026.01 | 32.74 | — | — | |
| DeFMBackbone=ViT-L/142026.01 | 31.26 | — | — | |
| B-SegNetdata=RGB2019.06 | 30.7 | 71.2 | 45.9 | |
| C-RADIOv3Backbone=ViT-L/16 (CPE)2026.01 | 28.17 | — | — | |
| DeFMBackbone=ViT-S/142026.01 | 27.78 | — | — | |
| DINOv2Backbone=ViT-L/142026.01 | 27.11 | — | — | |
| DeFMBackbone=ResNet-502026.01 | 26.02 | — | — | |
| DeFMBackbone=ResNet-342026.01 | 24.21 | — | — | |
| DeFMBackbone=ResNet-182026.01 | 19.85 | — | — | |
| DINOv2Backbone=ViT-S/142026.01 | 18.46 | — | — | |
| DINOv3Backbone=ViT-S/162026.01 | 18.42 | — | — | |
| TheiaBackbone=DeiT-S/162026.01 | 11.18 | — | — | |
| Cheng et al.data=RGBD2019.06 | — | — | 58 | |
| LSTMdata=RGBD2019.06 | — | — | 48.1 |