Object Discovery on COCO
0.413FG-ARISmoothSAi
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| SmoothSAi#slot=7, Input resolution=224x224, Encoder=DINO2 ViT-S/142025.08 | 0.413 | 0.334 | 0.293 | 0.318 | |
| DIAS#slot=7, Input resolution=224x224, Encoder=DINO2 ViT-S/142025.08 | 0.412 | 0.317 | 0.256 | 0.302 | |
| DINOSAUR-MLParchitecture=MLP2024.07 | 0.406 | — | — | — | |
| SPOT#slot=7, Input resolution=224x224, Encoder=DINO2 ViT-S/142025.08 | 0.404 | 0.309 | 0.237 | 0.293 | |
| SlotDiffusion2024.11 | 0.372 | 0.31 | — | — | |
| SlotDiffusion*source=results from [39]2024.07 | 0.372 | — | — | — | |
| DINOSAUR#slot=7, Input resolution=224x224, Encoder=DINO2 ViT-S/142025.08 | 0.37 | 0.283 | 0.182 | 0.269 | |
| SPOT2024.07 | 0.365 | — | — | — | |
| DINOSAUR-Trans.architecture=Transformer2024.07 | 0.352 | — | — | — | |
| LSD2024.11 | 0.35 | 0.304 | — | — | |
| GLASSvariant=Standard2024.07 | 0.341 | — | — | — | |
| SLATE2024.11 | 0.325 | 0.291 | — | — | |
| SLATE*source=results from [39]2024.07 | 0.325 | — | — | — | |
| GLASS+variant=GLASS+2024.07 | 0.325 | — | — | — | |
| SlotDiffusion#slot=7, Input resolution=224x224, Encoder=DINO2 ViT-S/142025.08 | 0.29 | 0.27 | 0.177 | 0.256 | |
| StableLSD2024.07 | 0.289 | — | — | — | |
| SLATE#slot=7, Input resolution=224x224, Encoder=DINO2 ViT-S/142025.08 | 0.288 | 0.268 | 0.175 | 0.254 | |
| Slot Attention2024.11 | 0.214 | 0.172 | — | — | |
| SA*source=results from [39]2024.07 | 0.214 | — | — | — | |
| AdaSlotDecoder=MLP, K-type=adaptive2026.01 | — | 0.274 | — | — | |
| MetaSlotDecoder=MLP, K-type=adaptive2026.01 | — | 0.295 | — | — | |
| QASADecoder=Trans., K-type=adaptive2026.01 | — | 0.367 | — | — | |
| QASA-MLPDecoder=MLP, K-type=adaptive2026.01 | — | 0.336 | — | — |