Point Tracking on TAP-Vid-Kinetics (val)
63.5Average Displacement ErrorPIPs++
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| PIPs++Train data=PO, Track Type=Sparse2026.02 | 63.5 | — | — | |
| DOTTrain data=Kub, Track Type=Dense2026.02 | 63.8 | 48.4 | 85.2 | |
| TAPIRTrain data=Kub, Track Type=Sparse2026.02 | 64.2 | 49.6 | 85 | |
| CoTrackerResolution=256x2562024.02 | 64.3 | 48.7 | 86.5 | |
| CoTracker2025.03 | 64.3 | 49.6 | 83.3 | |
| CoTrackerTrain data=Kub, Track Type=Sparse2026.02 | 64.3 | 49.6 | 83.3 | |
| TAPTR2025.03 | 64.4 | 49 | 85.2 | |
| TAPTRTrain data=Kub, Track Type=Sparse2026.02 | 64.4 | 49 | 85.2 | |
| Tuned TAPIRfix=true, snap=true, Resolution=256x2562024.02 | 65.3 | 52.5 | 85.5 | |
| Tuned TAPIRfix=true, Resolution=256x2562024.02 | 66 | 53.3 | 85.1 | |
| DELTATrain data=Kub, Track Type=Dense2026.02 | 66.6 | 51 | 84 | |
| LocoTrack2025.03 | 66.8 | 52.9 | 85.3 | |
| LocoTrackTrain data=Kub, Track Type=Sparse2026.02 | 66.8 | 52.9 | 85.3 | |
| CoTracker3Train data=Kub+15k, Track Type=Sparse, Inference Mode=offline2026.02 | 67.8 | 54.7 | 87.4 | |
| BootsTAPIRResolution=256x2562024.02 | 68.4 | 54.6 | 86.5 | |
| BootsTAPIR2025.03 | 68.4 | 54.6 | 86.5 | |
| BootsTAPIRTrain data=Kub+15M, Track Type=Sparse2026.02 | 68.4 | 54.6 | 86.5 | |
| BootsTAPIRfix=true, Resolution=256x2562024.02 | 68.5 | 54.7 | 86.3 | |
| CoTracker3Train data=Kub+15k, Track Type=Sparse, Inference Mode=online2026.02 | 68.5 | 55.8 | 88.3 | |
| BootsTAPIRfix=true, snap=true, Resolution=256x2562024.02 | 68.8 | 54.5 | 86.3 | |
| BootsTAPIRfix=true, snap=true, extra data=true, Resolution=256x2562024.02 | 68.8 | 55.8 | 86.6 | |
| CoTracker + OursBackbone=VGGT, Fine-tuned=Kubric2025.03 | 69 | 57.2 | 88.9 | |
| AllTrackerTrain data=Kub, Track Type=Dense2026.02 | 71.3 | 59.3 | 89.3 | |
| AllTrackerTrain data=Kub+Mix, Track Type=Dense2026.02 | 72.3 | 59.1 | 90.3 | |
| CoWTrackerTrain data=Kub, Track Type=Dense2026.02 | 73.1 | 60.9 | 91.5 |