Semantic Segmentation on Cityscapes (mIoU, mFSc)
82.98mIoUDepth Anything
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Depth AnythingPublication=CVPR’24, Type=Vision Foundation Models2026.02 | 82.98 | 90.31 | |
| DINO V2Publication=ICLR’23, Type=Vision Foundation Models2026.02 | 82.49 | 89.99 | |
| TwInS (Ours)Publication=/, Type=Joint Learning Approaches2026.02 | 82.13 | 89.73 | |
| RoadFormer+Publication=TIV’24, Type=Feature-Fusion Networks2026.02 | 80.73 | 88.83 | |
| Vit-CoMerPublication=CVPR’24, Type=Vision Foundation Models2026.02 | 79.72 | 88.25 | |
| Mask2FormerPublication=CVPR’22, Type=Single-Modal Networks2026.02 | 78.41 | 86.29 | |
| DFormerPublication=ICLR’24, Type=Feature-Fusion Networks2026.02 | 77.76 | 86.89 | |
| RoadFormerPublication=TIV’24, Type=Feature-Fusion Networks2026.02 | 76.82 | 86.11 | |
| DFormerv2Publication=CVPR’25, Type=Feature-Fusion Networks2026.02 | 75.56 | 85.36 | |
| CMXPublication=TITS’23, Type=Feature-Fusion Networks2026.02 | 73.55 | 83.76 | |
| SegmenterPublication=ICCV’21, Type=Single-Modal Networks2026.02 | 72.88 | 83.44 | |
| KNetPublication=NeurIPS’21, Type=Single-Modal Networks2026.02 | 72.78 | 83.08 | |
| TiCoSSPublication=TASE’25, Type=Joint Learning Approaches2026.02 | 63.57 | 74.06 | |
| SegFormerPublication=NeurIPS’21, Type=Single-Modal Networks2026.02 | 60.13 | 85.6 | |
| SemStereoPublication=AAAI’25, Type=Joint Learning Approaches2026.02 | 59.86 | 68.42 | |
| S3M-NetPublication=TIV’24, Type=Joint Learning Approaches2026.02 | 58.29 | 65.45 | |
| SG-RoadSegPublication=ICRA’24, Type=Joint Learning Approaches2026.02 | 56.37 | 68.47 | |
| SGDepthPublication=ECCV’20, Type=Joint Learning Approaches2026.02 | 55.96 | 61.11 | |
| DSNetPublication=ICRA’19, Type=Joint Learning Approaches2026.02 | 54.3 | 59.96 | |
| BiSeNet V2Publication=IJCV’21, Type=Single-Modal Networks2026.02 | 51.72 | 62.91 |