Point Tracking on TAP-Vid Kinetics (First)
22.4Avg Displacement Error (delta_avg)DeiT
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DeiTSetup=ZS, Size=32×32, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 22.4 | — | — | |
| DINOv2+NNSupervision=Zero-Shot2025.10 | 24.22 | 12.69 | 62.45 | |
| CLIPSetup=ZS, Size=32×32, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 25 | — | — | |
| DIFTSupervision=Zero-Shot2025.10 | 25.56 | 15.1 | 63.17 | |
| MAESetup=ZS, Size=32×32, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 27.6 | — | — | |
| SD-DINOSupervision=Zero-Shot2025.10 | 28.37 | 16.47 | 62.79 | |
| SAMSetup=WS, Size=32×32, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 31.4 | — | — | |
| SimVOS-B/16Setup=WS, Size=32×32, #L.P.=-, Data=D + Y (3.4k), Evaluation Mode=queried-first mode2026.03 | 32 | — | — | |
| DINOv2-S/14Setup=ZS, Size=32×32, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 33.4 | — | — | |
| DINOv2-B/14-RegSetup=ZS, Size=32×32, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 33.6 | — | — | |
| DINOSetup=ZS, Size=32×32, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 34.4 | — | — | |
| DINOv2-B/14Setup=ZS, Size=32×32, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 34.5 | — | — | |
| SMTCSetup=ZS, Size=30×45, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 36.1 | — | — | |
| DINOv3-S/16Setup=ZS, Size=32×32, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 36.7 | — | — | |
| SDSetup=ZS, Size=32×32, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 37.2 | — | — | |
| DINOv3-B/16Setup=ZS, Size=32×32, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 37.4 | — | — | |
| CRW (ResNet-18)Setup=ZS, Size=30×45, #L.P.=-, Data=V, Evaluation Mode=queried-first mode2026.03 | 40.9 | — | — | |
| Point PromptingSupervision=Zero-Shot2025.10 | 41.51 | 27.36 | 71.39 | |
| GMRWSupervision=Self-Sup.2025.10 | 41.63 | 25.7 | 71.33 | |
| DIFT (SD1.5)Setup=ZS, Size=30×45, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 41.9 | — | — | |
| VFS (ResNet-50)Setup=ZS, Size=30×45, #L.P.=-, Data=V, Evaluation Mode=queried-first mode2026.03 | 42.4 | — | — | |
| DIFT (SD2.1)Setup=ZS, Size=30×45, #L.P.=-, Data=S, Evaluation Mode=queried-first mode2026.03 | 42.9 | — | — | |
| DiffTrack (HunyuanVideo)Setup=ZS, Size=30×45, #L.P.=-, Data=V, Evaluation Mode=queried-first mode2026.03 | 45.5 | — | — | |
| M2Pv2-S/14-R@64Setup=WS, Size=32×32, #L.P.=4.7M, Data=D + Y (3.4k), Evaluation Mode=queried-first mode2026.03 | 45.5 | — | — | |
| DiffTrack (CogVideoX-2B)Setup=ZS, Size=30×45, #L.P.=-, Data=V, Evaluation Mode=queried-first mode2026.03 | 46.3 | — | — | |
| M2Pv2-B/14-R@64Setup=WS, Size=32×32, #L.P.=9.5M, Data=D + Y (3.4k), Evaluation Mode=queried-first mode2026.03 | 46.4 | — | — | |
| RAFTSupervision=Supervised2025.10 | 46.44 | 30.15 | 75.44 | |
| M2Pv3-S/16-R@64Setup=WS, Size=32×32, #L.P.=4.7M, Data=D + Y (3.4k), Evaluation Mode=queried-first mode2026.03 | 48.4 | — | — | |
| TAP-NetSupervision=Supervised2025.10 | 48.42 | 34.59 | 80.88 | |
| DiffTrack (CogVideoX-5B)Setup=ZS, Size=30×45, #L.P.=-, Data=V, Evaluation Mode=queried-first mode2026.03 | 49.2 | — | — | |
| Chrono-S/16Setup=FS, Size=32×32, #L.P.=16.2M, Data=Kubric (11k), Evaluation Mode=queried-first mode2026.03 | 49.3 | — | — | |
| M2Pv3-B/16-R@64Setup=WS, Size=32×32, #L.P.=9.5M, Data=D + Y (3.4k), Evaluation Mode=queried-first mode2026.03 | 49.4 | — | — | |
| TAPNetTraining Dataset=Kub, Evaluation Resolution=256 × 2562025.07 | 54.4 | 38.5 | 80.6 | |
| Opt-CWMSupervision=Self-Sup.2025.10 | 57.74 | 44.85 | 84.12 | |
| PIPs++2023.07 | 58.5 | — | — | |
| TAPIRSupervision=Supervised2025.10 | 59.56 | 47.46 | 85.76 | |
| TAPIR2023.07 | 64.2 | 49.6 | 85 | |
| TAPIRTraining Dataset=Kub, Evaluation Resolution=256 × 2562025.07 | 64.2 | 49.6 | 85 | |
| TAPTRv2Training Dataset=Kub, Evaluation Resolution=256 × 2562025.07 | 64.2 | 49.7 | 85.7 | |
| TAPTRTraining Dataset=Kub, Evaluation Resolution=256 × 2562025.07 | 64.4 | 49 | 85.2 | |
| TAPNextSupervision=Supervised2025.10 | 64.46 | 52.97 | 89.3 | |
| CoTracker2023.07 | 64.5 | 48.8 | 85.8 | |
| CoTracker2Training Dataset=Kub, Evaluation Resolution=384 × 5122025.07 | 64.5 | 48.8 | 85.8 | |
| CoTracker3Supervision=Supervised2025.10 | 65.99 | 54.35 | 89.43 | |
| LocoTrackTraining Dataset=Kub, Evaluation Resolution=384 × 5122025.07 | 66.4 | 52.3 | 82.1 | |
| LocoTrackTraining Dataset=Kub, Evaluation Resolution=256 × 2562025.07 | 66.8 | 52.9 | 85.3 | |
| Track-OnTraining Dataset=Kub, Evaluation Resolution=384 × 5122025.07 | 67.3 | 53.9 | 87.8 | |
| TAPTRv3Training Dataset=Kub, Evaluation Resolution=256 × 2562025.07 | 67.5 | 54.5 | 88.2 | |
| CoTracker3 (online)Training Dataset=Kub64+15K, Evaluation Resolution=384 × 5122025.07 | 67.8 | 54.7 | 87.4 | |
| BootsTAPIRTraining Dataset=Kub+15M, Evaluation Resolution=256 × 2562025.07 | 68.4 | 54.6 | 86.5 | |
| Anthro-LocoTrackTraining Dataset=Kub+1.4K, Evaluation Resolution=256 × 2562025.07 | 68.4 | 53.9 | 86.4 | |
| CoTracker3 (offline)Training Dataset=Kub64+15K, Evaluation Resolution=384 × 5122025.07 | 68.5 | 55.8 | 88.3 | |
| Anthro-LocoTrackTraining Dataset=Kub+1.4K, Evaluation Resolution=384 × 5122025.07 | 68.6 | 54.8 | 85.3 | |
| Online TAPIRTraining Dataset=Kub, Evaluation Resolution=256 × 2562025.07 | — | 51.5 | — |