Video Object Segmentation on DAVIS (J/F/T Metrics)
90.04J & F MeanFull Precision
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Full PrecisionModel=SAM2-L, Precision=FP2026.05 | 90.04 | — | — | — | — | — | — | — | |
| Full PrecisionModel=SAM2-B+, Precision=FP2026.05 | 89.75 | — | — | — | — | — | — | — | |
| Full PrecisionModel=SAM2-S, Precision=FP2026.05 | 89.42 | — | — | — | — | — | — | — | |
| Full PrecisionModel=SAM2-T, Precision=FP2026.05 | 89.1 | — | — | — | — | — | — | — | |
| CAR-SAMModel=SAM2-B+, Precision=W6A62026.05 | 89 | — | — | — | — | — | — | — | |
| QDropModel=SAM2-S, Precision=W6A62026.05 | 88.53 | — | — | — | — | — | — | — | |
| PTQ4SAMModel=SAM2-B+, Precision=W6A62026.05 | 88.51 | — | — | — | — | — | — | — | |
| CAR-SAMModel=SAM2-S, Precision=W6A62026.05 | 88.41 | — | — | — | — | — | — | — | |
| QDropModel=SAM2-B+, Precision=W6A62026.05 | 87.95 | — | — | — | — | — | — | — | |
| PTQ4SAMModel=SAM2-S, Precision=W6A62026.05 | 87.87 | — | — | — | — | — | — | — | |
| BRECQModel=SAM2-B+, Precision=W6A62026.05 | 87.84 | — | — | — | — | — | — | — | |
| CAR-SAMModel=SAM2-T, Precision=W6A62026.05 | 87.64 | — | — | — | — | — | — | — | |
| PTQ4SAMModel=SAM2-T, Precision=W6A62026.05 | 87.63 | — | — | — | — | — | — | — | |
| BRECQModel=SAM2-S, Precision=W6A62026.05 | 87.54 | — | — | — | — | — | — | — | |
| QDropModel=SAM2-T, Precision=W6A62026.05 | 87.45 | — | — | — | — | — | — | — | |
| BRECQModel=SAM2-T, Precision=W6A62026.05 | 87.22 | — | — | — | — | — | — | — | |
| CAR-SAMModel=SAM2-T, Precision=W4A42026.05 | 84.95 | — | — | — | — | — | — | — | |
| PTQ4SAMModel=SAM2-T, Precision=W4A42026.05 | 84.83 | — | — | — | — | — | — | — | |
| CAR-SAMModel=SAM2-S, Precision=W4A42026.05 | 84.79 | — | — | — | — | — | — | — | |
| QDropModel=SAM2-T, Precision=W4A42026.05 | 84.6 | — | — | — | — | — | — | — | |
| QDropModel=SAM2-S, Precision=W4A42026.05 | 84.52 | — | — | — | — | — | — | — | |
| PTQ4SAMModel=SAM2-S, Precision=W4A42026.05 | 84.32 | — | — | — | — | — | — | — | |
| BRECQModel=SAM2-S, Precision=W4A42026.05 | 84.12 | — | — | — | — | — | — | — | |
| BRECQModel=SAM2-T, Precision=W4A42026.05 | 84.1 | — | — | — | — | — | — | — | |
| CAR-SAMModel=SAM2-L, Precision=W6A62026.05 | 78.1 | — | — | — | — | — | — | — | |
| QDropModel=SAM2-L, Precision=W6A62026.05 | 77.87 | — | — | — | — | — | — | — | |
| PTQ4SAMModel=SAM2-L, Precision=W6A62026.05 | 77.55 | — | — | — | — | — | — | — | |
| BRECQModel=SAM2-L, Precision=W6A62026.05 | 76.12 | — | — | — | — | — | — | — | |
| Ours (WRC)Resolution=784, Upsampling Factor=2x2026.05 | 74.39 | — | — | — | — | — | — | — | |
| Ours (WRC)Resolution=672, Upsampling Factor=2x2026.05 | 73.57 | — | — | — | — | — | — | — | |
| AnyUpResolution=784, Upsampling Factor=2x2026.05 | 72.44 | — | — | — | — | — | — | — | |
| PTQ4SAMModel=SAM2-B+, Precision=W4A42026.05 | 71.89 | — | — | — | — | — | — | — | |
| RaysUp2026.06 | 71.47 | 68.14 | — | — | 74.81 | — | — | — | |
| Ours (WRC)Resolution=560, Upsampling Factor=2x2026.05 | 71.35 | — | — | — | — | — | — | — | |
| QDropModel=SAM2-B+, Precision=W4A42026.05 | 71.3 | — | — | — | — | — | — | — | |
| AnyUpResolution=672, Upsampling Factor=2x2026.05 | 71.21 | — | — | — | — | — | — | — | |
| AnyUp2026.06 | 70.98 | 67.55 | — | — | 74.42 | — | — | — | |
| LoftUp2026.06 | 70.92 | 67.32 | — | — | 74.53 | — | — | — | |
| JAFAR2026.06 | 70.9 | 67.18 | — | — | 74.62 | — | — | — | |
| BRECQModel=SAM2-B+, Precision=W4A42026.05 | 70.89 | — | — | — | — | — | — | — | |
| WRCInput Resolution=784, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 70.77 | — | — | — | — | — | — | — | |
| CAR-SAMModel=SAM2-B+, Precision=W4A42026.05 | 70.18 | — | — | — | — | — | — | — | |
| LiFTResolution=784, Upsampling Factor=2x2026.05 | 69.78 | — | — | — | — | — | — | — | |
| DINOv3Input Resolution=784, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 69.64 | — | — | — | — | — | — | — | |
| Ours (WRC)Resolution=504, Upsampling Factor=2x2026.05 | 69.28 | — | — | — | — | — | — | — | |
| WRCInput Resolution=672, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 69.1 | — | — | — | — | — | — | — | |
| AnyUpResolution=560, Upsampling Factor=2x2026.05 | 69.03 | — | — | — | — | — | — | — | |
| LiFTResolution=672, Upsampling Factor=2x2026.05 | 68.86 | — | — | — | — | — | — | — | |
| FeatUp2026.06 | 68.74 | 65.45 | — | — | 72.03 | — | — | — | |
| LiFTInput Resolution=784, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 68.24 | — | — | — | — | — | — | — | |
| JAFARInput Resolution=784, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 67.91 | — | — | — | — | — | — | — | |
| DINOv3Input Resolution=672, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 67.89 | — | — | — | — | — | — | — | |
| DINOv2Resolution=784, Upsampling Factor=2x2026.05 | 67.53 | — | — | — | — | — | — | — | |
| AnyUpResolution=504, Upsampling Factor=2x2026.05 | 67.17 | — | — | — | — | — | — | — | |
| JAFARInput Resolution=672, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 67.12 | — | — | — | — | — | — | — | |
| Ours (WRC)Resolution=448, Upsampling Factor=2x2026.05 | 66.09 | — | — | — | — | — | — | — | |
| LiFTInput Resolution=672, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 66 | — | — | — | — | — | — | — | |
| LiFTResolution=560, Upsampling Factor=2x2026.05 | 65.76 | — | — | — | — | — | — | — | |
| AnyUpResolution=448, Upsampling Factor=2x2026.05 | 65.55 | — | — | — | — | — | — | — | |
| WRCInput Resolution=560, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 65.17 | — | — | — | — | — | — | — | |
| JAFARResolution=784, Upsampling Factor=2x2026.05 | 65.05 | — | — | — | — | — | — | — | |
| DINOv2Resolution=672, Upsampling Factor=2x2026.05 | 65.03 | — | — | — | — | — | — | — | |
| BilinearResolution=784, Upsampling Factor=2x2026.05 | 64.99 | — | — | — | — | — | — | — | |
| Bilinear2026.06 | 64.87 | 63.21 | — | — | 66.53 | — | — | — | |
| JAFARInput Resolution=560, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 64.61 | — | — | — | — | — | — | — | |
| AnyUpInput Resolution=784, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 64.2 | — | — | — | — | — | — | — | |
| JAFARResolution=560, Upsampling Factor=2x2026.05 | 64.01 | — | — | — | — | — | — | — | |
| DINOv3Input Resolution=560, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 63.79 | — | — | — | — | — | — | — | |
| JAFARResolution=672, Upsampling Factor=2x2026.05 | 63.59 | — | — | — | — | — | — | — | |
| LiFTResolution=504, Upsampling Factor=2x2026.05 | 63.06 | — | — | — | — | — | — | — | |
| JAFARResolution=504, Upsampling Factor=2x2026.05 | 63.06 | — | — | — | — | — | — | — | |
| BilinearResolution=672, Upsampling Factor=2x2026.05 | 62.52 | — | — | — | — | — | — | — | |
| BilinearInput Resolution=784, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 62.44 | — | — | — | — | — | — | — | |
| AnyUpInput Resolution=672, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 62.43 | — | — | — | — | — | — | — | |
| JAFARInput Resolution=448, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 61.31 | — | — | — | — | — | — | — | |
| WRCInput Resolution=448, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 61.3 | — | — | — | — | — | — | — | |
| LiFTInput Resolution=560, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 61.13 | — | — | — | — | — | — | — | |
| DINOv2Resolution=560, Upsampling Factor=2x2026.05 | 61.12 | — | — | — | — | — | — | — | |
| DINO+LiFTResolution=4482024.03 | 61.11 | 59.43 | — | — | 62.79 | — | — | — | |
| ToBoPre-trained=Kinetics-400, Epochs=4002025.07 | 60.6 | 58.4 | — | — | 63 | — | — | — | |
| AnyUpInput Resolution=560, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 60.45 | — | — | — | — | — | — | — | |
| JAFARResolution=448, Upsampling Factor=2x2026.05 | 60.37 | — | — | — | — | — | — | — | |
| RSPPre-trained=Kinetics-400, Epochs=4002025.07 | 60.1 | 57.4 | — | — | 62.8 | — | — | — | |
| DINOPre-trained=Kinetics-400, Epochs=4002025.07 | 59.5 | 56.5 | — | — | 62.5 | — | — | — | |
| BilinearResolution=560, Upsampling Factor=2x2026.05 | 59.37 | — | — | — | — | — | — | — | |
| BilinearInput Resolution=672, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 59.3 | — | — | — | — | — | — | — | |
| LiFTResolution=448, Upsampling Factor=2x2026.05 | 59.25 | — | — | — | — | — | — | — | |
| CropMAEPre-trained=Kinetics-400, Epochs=4002025.07 | 58.6 | 55.8 | — | — | 61.4 | — | — | — | |
| C-JEPAPretrain Epochs=600, Backbone=ViT-L/162024.10 | 58.3 | — | — | — | — | — | — | — | |
| DINOv2Resolution=504, Upsampling Factor=2x2026.05 | 58.29 | — | — | — | — | — | — | — | |
| SiamMAEPre-trained=Kinetics-400, Epochs=4002025.07 | 58.1 | 56.6 | — | — | 59.6 | — | — | — | |
| MoCo v3Pre-trained=Kinetics-400, Epochs=4002025.07 | 57.7 | 54.6 | — | — | 60.8 | — | — | — | |
| AnyUpInput Resolution=448, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 57.57 | — | — | — | — | — | — | — | |
| LiFTInput Resolution=448, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 57.31 | — | — | — | — | — | — | — | |
| DINOv3Input Resolution=448, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 56.69 | — | — | — | — | — | — | — | |
| I-JEPAPretrain Epochs=600, Backbone=ViT-L/162024.10 | 56.6 | — | — | — | — | — | — | — | |
| BilinearResolution=504, Upsampling Factor=2x2026.05 | 55.67 | — | — | — | — | — | — | — | |
| BilinearInput Resolution=560, Backbone=DINOv3-ViT-S/16, Upsampling ratio=2x2026.05 | 54.92 | — | — | — | — | — | — | — | |
| DINO+JBUResolution=4482024.03 | 54.5 | 54.9 | — | — | 54.1 | — | — | — | |
| DINOv2Resolution=448, Upsampling Factor=2x2026.05 | 54.16 | — | — | — | — | — | — | — |