Camera Pose Estimation on ScanNet-50 (300 frames)
0.08ATEpi^3
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| pi^3Dense settings=true, Full-attention baseline=true, Spd.=1.0×, Backbone=pi^32026.07 | 0.08 | 2.16 | — | — | — | 0.02 | 0.52 | |
| Fast pi^3Dense settings=true, Full-attention baseline=false, Spd.=2.0×, Backbone=pi^32026.07 | 0.09 | 2.37 | — | — | — | 0.02 | 0.55 | |
| SAF3RDense settings=true, Full-attention baseline=false, Spd.=5.3×, Backbone=pi^32026.07 | 0.09 | 2.36 | — | — | — | 0.02 | 0.54 | |
| FastVGGT2026.06 | 0.093 | 4.04 | 0.776 | 0.036 | 17.8 | — | — | |
| RegimeVGGT2026.06 | 0.095 | 4.03 | 0.78 | 0.038 | 12.2 | — | — | |
| VGGT*Checkpointing=Activation checkpointing2026.06 | 0.096 | 4.206 | 0.786 | 0.04 | 48.7 | — | — | |
| A pi^3Dense settings=true, Full-attention baseline=false, Spd.=4.9×, Backbone=pi^32026.07 | 0.1 | 2.39 | — | — | — | 0.02 | 0.55 | |
| FastVGGTDense settings=true, Full-attention baseline=false, Spd.=2.8×, Backbone=VGGT2026.07 | 0.13 | 3.48 | — | — | — | 0.03 | 0.77 | |
| DA3Dense settings=true, Full-attention baseline=true, Spd.=1.0×, Backbone=DA32026.07 | 0.13 | 3.68 | — | — | — | 0.02 | 0.82 | |
| FastDA3Dense settings=true, Full-attention baseline=false, Spd.=2.7×, Backbone=DA32026.07 | 0.13 | 3.77 | — | — | — | 0.02 | 1 | |
| S-VGGT2026.06 | 0.137 | 4.287 | 1.307 | 0.122 | 14.3 | — | — | |
| VGGTDense settings=true, Full-attention baseline=true, Spd.=1.0×, Backbone=VGGT2026.07 | 0.14 | 3.64 | — | — | — | 0.03 | 0.74 | |
| Sparse pi^3Dense settings=true, Full-attention baseline=false, Spd.=2.7×, Backbone=pi^32026.07 | 0.14 | 4.99 | — | — | — | 0.02 | 0.61 | |
| AVGGTDense settings=true, Full-attention baseline=false, Spd.=3.1×, Backbone=VGGT2026.07 | 0.15 | 3.95 | — | — | — | 0.04 | 0.97 | |
| SAF3RDense settings=true, Full-attention baseline=false, Spd.=3.2×, Backbone=VGGT2026.07 | 0.15 | 3.97 | — | — | — | 0.04 | 0.85 | |
| SAF3RDense settings=true, Full-attention baseline=false, Spd.=3.3×, Backbone=DA32026.07 | 0.15 | 5.19 | — | — | — | 0.03 | 1.4 | |
| SparseVGGTDense settings=true, Full-attention baseline=false, Spd.=2.5×, Backbone=VGGT2026.07 | 0.16 | 4.51 | — | — | — | 0.04 | 0.96 | |
| Co-MeDense settings=true, Full-attention baseline=false, Spd.=2.6×, Backbone=VGGT2026.07 | 0.16 | 3.96 | — | — | — | 0.05 | 1.24 | |
| ADA3Dense settings=true, Full-attention baseline=false, Spd.=3.2×, Backbone=DA32026.07 | 0.35 | 12.43 | — | — | — | 0.11 | 7.34 |