Visual Object Tracking on NfS (Pr, SR, VNE)
71.1Success Rate (SR)GOT-Edit-378
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GOT-Edit-378Semantic Feature=DINOv2-L, Geometry Feature=VGGT, Training Data of Tracker=Base+VastTrack, Frames=3, Trainable Parameters=53M2026.07 | 71.1 | — | — | |
| GOT-Edit-378Semantic Feature=DINOv2-L, Geometry Feature=DA3-L, Training Data of Tracker=Base+VastTrack, Frames=3, Trainable Parameters=28M2026.07 | 70.8 | — | — | |
| GOT-JEPAVariant=Highest-resolution (ViT-L prioritized)2026.07 | 70.8 | — | — | |
| GOT-Edit-252Semantic Feature=DINOv2-L, Geometry Feature=DA3-L, Training Data of Tracker=Base+VastTrack, Frames=3, Trainable Parameters=51M2026.07 | 70.7 | — | — | |
| MCITrack-384Semantic Feature=Fast-iTPN-L, Training Data of Tracker=Base+VastTrack, Frames=5, Trainable Parameters=287M2026.07 | 70.6 | — | — | |
| GOT-Edit-378Semantic Feature=DINOv2-L, Geometry Feature=VGGT, Training Data of Tracker=Base, Frames=3, Trainable Parameters=53M2026.07 | 69.9 | — | — | |
| SAMURAIYear=2026, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 69.2 | — | — | |
| ToMP-378 (Reproduced)Semantic Feature=DINOv2-L, Training Data of Tracker=Base+VastTrack, Frames=3, Trainable Parameters=25M2026.07 | 69 | — | — | |
| ARTrackv2Year=2024, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 68.4 | — | — | |
| PiVOT-378Semantic Feature=DINOv2-L, Training Data of Tracker=Base, Frames=3, Trainable Parameters=34M2026.07 | 68.2 | — | — | |
| PiVOTYear=2025, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 68.2 | — | — | |
| HIPTrackYear=2024, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 68.2 | — | — | |
| ToMP-378Semantic Feature=DINOv2-L, Training Data of Tracker=Base, Frames=3, Trainable Parameters=25M2026.07 | 67.8 | — | — | |
| SuperSBTYear=2024, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 67.7 | — | — | |
| UniSOTYear=2026, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 67.6 | — | — | |
| SATrackYear=2025, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 67.5 | — | — | |
| ROMTrackYear=2023, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 67.5 | — | — | |
| ToMPYear=2022, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 67 | — | — | |
| GRM-320Semantic Feature=ViT-MAE-L, Training Data of Tracker=Base, Frames=32026.07 | 66.9 | — | — | |
| MPITYear=2025, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 66.9 | — | — | |
| GRMYear=2023, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 66.9 | — | — | |
| MFDSTrackYear=2025, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 66.8 | — | — | |
| LoRAT-378Semantic Feature=DINOv2-L, Training Data of Tracker=Base, Frames=3, Trainable Parameters=32M2026.07 | 66.7 | — | — | |
| SeqTrack-384Semantic Feature=ViT-MAE-L, Training Data of Tracker=Base, Frames=3, Trainable Parameters=309M2026.07 | 66.7 | — | — | |
| LoRATYear=2024, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 66.7 | — | — | |
| OSTrackYear=2022, Variant=Highest-resolution (ViT-L prioritized)2026.07 | 66.5 | — | — | |
| HTNN2026.06 | 65.5 | 64.8 | 2.81 | |
| ANNSiamFC2026.06 | 65.1 | 63.7 | 3.25 | |
| HSTNNType=Neuron-scale hybridization baseline2026.06 | 61.9 | 58.9 | 4.2 | |
| DeTrack-384Semantic Feature=Denoising ViT, Training Data of Tracker=Base, Frames=32026.07 | 60.2 | — | — | |
| SATrack-384Semantic Feature=SAViT, Training Data of Tracker=Base, Frames=62026.07 | 58.4 | — | — | |
| CANNDHType=Direct ANN-CANN hybridization baseline2026.06 | 42.7 | 38.7 | 2.84 | |
| FlyNetType=Heuristic ANN-CANN hybridization baseline2026.06 | 4 | 5.5 | 2.98 | |
| CANNMType=Motion-driven CANN baseline2026.06 | 3.4 | 3.9 | 3.17 |