Panoptic Segmentation on COCO
58PQHIPIE
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| HIPIEBackbone=ViT-H2023.07 | 58 | — | — | |
| HIPIEBackbone=ViT-H2023.07 | 58 | — | — | |
| Full-precisionBits (W/A)=16/162026.06 | 57.9 | — | — | |
| Mask2Former2022.12 | 57.8 | — | — | |
| Shift-and-Sum Quantization + LiteVARBits (W/A)=6/62026.06 | 57.8 | — | — | |
| Shift-and-Sum QuantizationBits (W/A)=6/62026.06 | 57.6 | — | — | |
| LiteVARBits (W/A)=6/62026.06 | 57.4 | — | — | |
| Ours-D3S2Backbone Scale=Base, Composition=D3 (finetuned master) + S2 (frozen auxiliary)2026.05 | 56.9 | — | — | |
| GELLA-13BEpoch=502024.02 | 56.5 | 83.3 | 67.3 | |
| GELLA-7BEpoch=502024.02 | 56.4 | 83.4 | 66.8 | |
| X-DecoderBackbone=DaViT-B2023.07 | 56.2 | — | — | |
| SEEMBackbone=DaViT-B2023.07 | 56.2 | — | — | |
| X-DecoderBackbone=DaViT-B2023.07 | 56.2 | — | — | |
| SEEMBackbone=DaViT-B2023.07 | 56.2 | — | — | |
| BRECQBits (W/A)=6/62026.06 | 55.8 | — | — | |
| DINOv3Backbone Scale=Base, Status=finetuned master2026.05 | 55.6 | — | — | |
| ODISEBackbone=ViT-H+SD2023.07 | 55.4 | — | — | |
| ODISEBackbone=ViT-H+SD2023.07 | 55.4 | — | — | |
| ODISE2026.03 | 55.4 | — | — | |
| OVRCOAT2026.03 | 54.6 | 82.9 | 65.1 | |
| FC-CLIP2026.03 | 54.4 | — | — | |
| Mask2FormerVenue=CVPR2022, Epoch=502024.02 | 53.4 | 83.1 | 63.4 | |
| GELLA-13BEpoch=122024.02 | 53.2 | 83.2 | 63.9 | |
| GELLA-7BEpoch=122024.02 | 53.1 | 83.3 | 63.8 | |
| HIPIEBackbone=RN502023.07 | 52.7 | — | — | |
| HIPIEBackbone=RN502023.07 | 52.7 | — | — | |
| X-DecoderBackbone=FocalT2023.07 | 52.6 | — | — | |
| X-DecoderBackbone=FocalT2023.07 | 52.6 | — | — | |
| OPSNet2026.03 | 52.4 | 83.5 | 62.1 | |
| Shift-and-Sum Quantization + LiteVARBits (W/A)=4/42026.06 | 52.2 | — | — | |
| Shift-and-Sum QuantizationBits (W/A)=4/42026.06 | 51.6 | — | — | |
| LiteVARBits (W/A)=4/42026.06 | 51.5 | — | — | |
| BRECQBits (W/A)=4/42026.06 | 51 | — | — | |
| SEEMBackbone=FocalT2023.07 | 50.6 | — | — | |
| SEEMBackbone=FocalT2023.07 | 50.6 | — | — | |
| MAFT+pan2026.03 | 50.3 | 82.2 | 60.3 | |
| Mask2FormerVenue=CVPR2022, Epoch=122024.02 | 49.2 | 82.5 | 58.8 | |
| UViMBackbone=ViT-L2022.12 | 45.8 | — | — | |
| DETR2022.12 | 45.6 | — | — | |
| SAM2Backbone Scale=Base, Status=frozen auxiliary2026.05 | 43.7 | — | — | |
| PainterBackbone=ViT-L2022.12 | 43.4 | — | — | |
| DeCon-ML-LPret. Dataset=COCO, Backbone=ResNet-50, Pret. Dec.=FPN2025.03 | 40.9 | — | — | |
| DeCon-ML-SPret. Dataset=COCO, Backbone=ResNet-50, Pret. Dec.=FPN2025.03 | 40.52 | — | — | |
| DeCon-SLPret. Dataset=COCO, Backbone=ResNet-50, Pret. Dec.=FCN2025.03 | 40.44 | — | — | |
| SlotCon (2022)Pret. Dataset=COCO, Backbone=ResNet-50, Pret. Dec.=-2025.03 | 40.31 | — | — | |
| Random init.Pret. Dataset=-, Backbone=ResNet-50, Pret. Dec.=-2025.03 | 33.51 | — | — |