Object-Centric Learning on MOVi-C
49MBO^iSPOT+FS+RC
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SPOT+FS+RCDec.=Trans., K-type=fixed2026.01 | 49 | 47.8 | |
| SPOT+FS+RCCategory=Autoregressive Teacher-Student2026.03 | 49 | 47.8 | |
| Slot Merging2026.03 | 48.35 | 47.43 | |
| SPOTEnsemble=true, Seeds=32023.12 | 47.3 | 46.7 | |
| SPOTDec.=Trans., K-type=fixed2026.01 | 47.3 | 46.7 | |
| SPOTCategory=Autoregressive Teacher-Student2026.03 | 47.3 | 46.7 | |
| SPOTEnsemble=false, Seeds=32023.12 | 47 | 46.4 | |
| QASADec.=Trans., K-type=adaptive2026.01 | 46.9 | 46.1 | |
| Top-Down Slot Attention2026.03 | 46.8 | 45.9 | |
| DINOSAUR+FS+RCDec.=MLP, K-type=fixed2026.01 | 46 | 44.8 | |
| DINOSAUR+FS+RCCategory=DINOSAUR-Based2026.03 | 46 | 44.8 | |
| Stable-LSD2023.12 | 45.6 | 44.2 | |
| LSDCategory=Diffusion2026.03 | 45.6 | 44.2 | |
| DINOSAUREncoder=ViT-B/16, Decoder=Autoregressive2023.12 | 42.4 | — | |
| DINOSAURDec.=Trans., K-type=fixed2026.01 | 42.4 | — | |
| DINOSAURCategory=DINOSAUR-Based2026.03 | 42.4 | — | |
| QASA-MLPDec.=MLP, K-type=adaptive2026.01 | 42.2 | 41.1 | |
| SLATE2023.12 | 39.4 | 37.8 | |
| SLATE2026.01 | 39.4 | 37.8 | |
| SLATE2026.03 | 39.4 | 37.8 | |
| DINOSAUR-MLPEncoder=ViT-S/8, Decoder=MLP2023.12 | 39.1 | — | |
| DINOSAUR-MLPDec.=MLP, K-type=fixed2026.01 | 39.1 | — | |
| DINOSAUR-MLPCategory=DINOSAUR-Based2026.03 | 39.1 | — | |
| AdSlotDec.=MLP, K-type=adaptive2026.01 | 35.6 | — | |
| AdaSlotCategory=Adaptive Slot Attention2026.03 | 35.6 | — | |
| MetaSlotDec.=MLP, K-type=adaptive2026.01 | 35 | — | |
| MetaSlot-MLPCategory=Adaptive Slot Attention2026.03 | 35 | — | |
| SA2023.12 | 26.2 | — | |
| SA2026.03 | 26.2 | — | |
| SLASH2023.12 | — | 27.7 | |
| SLASH2026.03 | — | 27.7 |