Semantic segmentation assigns a class label to every pixel — the dense prediction problem that underpins autonomous driving, medical imaging, and satellite analysis. FCN (2015) showed you could repurpose classifiers for pixel labeling, DeepLab introduced atrous convolutions and CRFs, and SegFormer (2021) proved transformers dominate here too. State-of-the-art on Cityscapes exceeds 85 mIoU, but ADE20K with its 150 classes remains brutally challenging. The frontier has moved toward universal segmentation models like Mask2Former that handle semantic, instance, and panoptic segmentation in a single architecture.
20K training, 2K validation images annotated with 150 object categories. Complex scene parsing benchmark.
Leading models on ADE20K.
| # | Model | mIoU | Year | Source |
|---|---|---|---|---|
| ★ | InternImage-H | 62.9 | 2025 | paper ↗ |
| 2 | BEiT-3 (ViT-L) | 62.8 | 2026 | paper ↗ |
| 3 | DINOv3 + Mask2Former (simple) | 62.6 | 2025 | paper ↗ |
| 4 | DINOv2 (ViT-g) + Linear | 62.0 | 2026 | paper ↗ |
| 5 | EoMT (ViT-L) | 58.4 | 2025 | paper ↗ |
| 6 | BEiT-L+ | 57.9 | 2021 | paper ↗ |
| 7 | Mask2Former (Swin-L) | 57.3 | 2025 | paper ↗ |
| 8 | Mask2Former (Swin-L) | 57.3 | 2026 | paper ↗ |
| 9 | OneFormer (Swin-L) | 57.0 | 2022 | paper ↗ |
| 10 | Mask2Former + Swin-L-FaPN | 56.4 | 2021 | paper ↗ |
Didn't find the model, metric, or dataset you needed? Tell us in one line. We read every message and reply within 48 hours.
2 datasets tracked for this task.
Still looking for something on Semantic Segmentation? A missing model, a stale score, a benchmark we should cover — drop it here and we'll handle it.
Real humans read every message. We track what people are asking for and prioritize accordingly.