Camera Pose Estimation on Sintel
0.061ATEπ³+Ours
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| π³+OursStream=true2025.12 | 0.061 | 0.028 | 0.249 | — | |
| π³Stream=false2025.12 | 0.073 | 0.037 | 0.287 | — | |
| π^3Inference Speed Complexity=O(N^2)2026.03 | 0.073 | 0.038 | 0.288 | — | |
| π³2026.02 | 0.074 | 0.04 | 0.282 | — | |
| pi^32026.01 | 0.074 | 0.04 | 0.282 | — | |
| Pi32026.03 | 0.074 | — | 0.282 | 0.04 | |
| pi^32025.07 | 0.074 | 0.04 | 0.282 | — | |
| AnyCamintrinsics_required=no, no_supervision=true, approx_runtime=<2min, test_time_refinement=true2025.03 | 0.078 | 0.031 | 0.427 | — | |
| MoRe(FA)type=FA2026.03 | 0.0877 | 0.058 | 0.3899 | — | |
| TrackingWorld (DELTA)Category=Joint depth & pose, 2D tracking source=DELTA [4]2025.12 | 0.088 | 0.035 | 0.41 | — | |
| LeapVOintrinsics_required=yes, no_supervision=false, approx_runtime=<2min2025.03 | 0.089 | 0.066 | 1.25 | — | |
| LEAP-VOCategory=I (SLAM-based methods requiring intrinsics)2026.01 | 0.089 | 0.066 | 1.25 | — | |
| Ours w/o ref.No sup.=true, Time=< 20s, Zero-shot=true2026.04 | 0.09 | 0.043 | 0.601 | — | |
| SirenPose2025.12 | 0.097 | 0.03 | 0.437 | — | |
| AnyCam (w/o ref.)intrinsics_required=no, no_supervision=true, approx_runtime=<20sec, test_time_refinement=false2025.03 | 0.099 | 0.045 | 0.567 | — | |
| AnyCam w/o ref.No sup.=true, Time=< 20s, Zero-shot=true2026.04 | 0.099 | 0.045 | 0.567 | — | |
| TrackingWorld (CoTrackerV3)Category=Joint depth & pose, 2D tracking source=CoTrackerV3 [1]2025.12 | 0.103 | 0.039 | 0.439 | — | |
| V-DPM2026.01 | 0.105 | 0.048 | 0.67 | — | |
| MonST3R+intrinsics_required=no, no_supervision=true, approx_runtime=<2min2025.03 | 0.108 | 0.042 | 0.732 | — | |
| TrajVG2026.02 | 0.108 | 0.038 | 0.274 | — | |
| MonST3R2026.01 | 0.108 | 0.042 | 0.732 | — | |
| MonST3ROptim.=•2025.10 | 0.108 | 0.042 | 0.732 | — | |
| Easi3RType=FA2026.01 | 0.11 | 0.042 | 0.758 | — | |
| Easi3RType=Optim2025.12 | 0.11 | 0.042 | 0.758 | — | |
| Easi3R2026.03 | 0.11 | 0.042 | 0.758 | — | |
| MonST3RCategory=Joint depth & pose2024.12 | 0.111 | 0.044 | 0.78 | — | |
| MonST3R-GAType=Optim2026.01 | 0.111 | 0.044 | 0.869 | — | |
| MonST3R-GACategory=II (Optimization-based VFM methods)2026.01 | 0.111 | 0.044 | 0.869 | — | |
| MonST3RCategory=Joint depth & pose2025.12 | 0.111 | 0.044 | 0.78 | — | |
| MonST3RType=Optim2025.12 | 0.111 | 0.044 | 0.869 | — | |
| MonST3R2026.03 | 0.111 | 0.044 | 0.869 | — | |
| MonST3R-GAType=Optim2026.03 | 0.111 | 0.044 | 0.869 | — | |
| MonST3R-GAType=Pair-wise2025.07 | 0.111 | 0.044 | 0.869 | — | |
| MonST3R-GAOptim.=true, Onl.=false2026.05 | 0.111 | 0.044 | 0.869 | — | |
| Monst3R2025.12 | 0.112 | 0.046 | 0.693 | — | |
| DPVOCategory=Pose only, ground truth camera intrinsics=true2024.12 | 0.115 | 0.072 | 1.975 | — | |
| DPVOintrinsics_required=yes, no_supervision=false, approx_runtime=<2min2025.03 | 0.115 | 0.072 | 1.975 | — | |
| DPVOCategory=I (SLAM-based methods requiring intrinsics)2026.01 | 0.115 | 0.072 | 1.975 | — | |
| DPVOCategory=Pose only, Ground truth camera intrinsics=true2025.12 | 0.115 | 0.072 | 1.975 | — | |
| Uni4DCategory=Joint depth & pose, 2D tracking source=reproduced by 2D tracks from DELTA2025.12 | 0.116 | 0.046 | 0.603 | — | |
| Align3R (Depth Pro)Category=Joint depth & pose, Monocular depth backbone=Depth Pro2024.12 | 0.128 | 0.042 | 0.432 | — | |
| Align3RCategory=Joint depth & pose, Depth Prior=Depth Pro [54]2025.12 | 0.128 | 0.042 | 0.432 | — | |
| ParticleSfMintrinsics_required=no, no_supervision=true, approx_runtime=<20min2025.03 | 0.129 | 0.031 | 0.525 | — | |
| VGGT+OursStream=true2025.12 | 0.131 | 0.053 | 0.398 | — | |
| DAGE2026.03 | 0.132 | — | 0.406 | 0.051 | |
| ZipMapInference Speed Complexity=O(N)2026.03 | 0.132 | 0.066 | 0.438 | — | |
| CasualSAMCategory=Joint depth & pose2024.12 | 0.141 | 0.035 | 0.615 | — | |
| CasualSAMintrinsics_required=no, no_supervision=true, approx_runtime=>1h2025.03 | 0.141 | 0.035 | 0.615 | — | |
| CasualSAMType=Optim2026.01 | 0.141 | 0.035 | 0.615 | — | |
| CasualSAM2026.01 | 0.141 | 0.035 | 0.615 | — | |
| CasualSAMCategory=II (Optimization-based VFM methods)2026.01 | 0.141 | 0.035 | 0.615 | — | |
| CasualSAMType=Pair-wise2025.07 | 0.141 | 0.035 | 0.615 | — | |
| CasualSAMOptim.=true, Onl.=false2026.05 | 0.141 | 0.035 | 0.615 | — | |
| PAGE-4D2025.10 | 0.143 | 0.078 | 0.538 | — | |
| MoRetype=Streaming2026.03 | 0.1474 | 0.0776 | 0.6157 | — | |
| Pi3Resolution=252px2026.03 | 0.153 | — | 0.684 | 0.088 | |
| ZipMapmode=streaming2026.03 | 0.1593 | 0.0655 | 0.7508 | — | |
| Align3R (Depth Anything V2)Category=Joint depth & pose, Monocular depth backbone=Depth Anything V22024.12 | 0.163 | 0.062 | 0.419 | — | |
| DPVO2025.12 | 0.163 | 0.057 | 1.196 | — | |
| VGGT2026.02 | 0.167 | 0.062 | 0.491 | — | |
| VGGT2026.03 | 0.167 | — | 0.491 | 0.062 | |
| VGGT2025.07 | 0.167 | 0.062 | 0.491 | — | |
| VGGT2026.03 | 0.168 | 0.065 | 0.477 | — | |
| VGGTType=Dense-view2025.07 | 0.169 | 0.064 | 0.474 | — | |
| VGGTStream=false2025.12 | 0.171 | 0.062 | 0.471 | — | |
| VGGTtype=FA2026.03 | 0.1715 | 0.0617 | 0.4695 | — | |
| VGGTType=FA2026.01 | 0.172 | 0.062 | 0.471 | — | |
| VGGTInference Speed Complexity=O(N^2)2026.03 | 0.172 | 0.061 | 0.471 | — | |
| VGGT2026.03 | 0.172 | 0.062 | 0.471 | — | |
| VGGTType=Offline2026.03 | 0.174 | 0.056 | 0.465 | — | |
| DROID-SLAMCategory=Pose only, ground truth camera intrinsics=true2024.12 | 0.175 | 0.084 | 1.912 | — | |
| DROID-SLAMintrinsics_required=yes, no_supervision=false, approx_runtime=<2min2025.03 | 0.175 | 0.084 | 1.912 | — | |
| DROID-SLAMCategory=I (SLAM-based methods requiring intrinsics)2026.01 | 0.175 | 0.084 | 1.912 | — | |
| DROID-SLAMCategory=Pose only, Ground truth camera intrinsics=true2025.12 | 0.175 | 0.084 | 1.912 | — | |
| CogniMap3DCategory=III (Feed-forward VFM methods)2026.01 | 0.176 | 0.068 | 0.6 | — | |
| MASt3R-GAType=Optim2026.01 | 0.185 | 0.06 | 1.496 | — | |
| MASt3R-GACategory=II (Optimization-based VFM methods)2026.01 | 0.185 | 0.06 | 1.496 | — | |
| MASt3RType=Optim2025.12 | 0.185 | 0.06 | 1.496 | — | |
| MASt3R2026.03 | 0.185 | 0.06 | 1.496 | — | |
| MASt3R-GAType=Optim2026.03 | 0.185 | 0.06 | 1.496 | — | |
| MASt3R-GAType=Pair-wise2025.07 | 0.185 | 0.06 | 1.496 | — | |
| MASt3R-GAOptim.=true, Onl.=false2026.05 | 0.185 | 0.06 | 1.496 | — | |
| Aether2026.02 | 0.189 | 0.054 | 0.694 | — | |
| VGGTCategory=III (Feed-forward VFM methods)2026.01 | 0.189 | 0.069 | 0.529 | — | |
| AETHER2026.03 | 0.189 | 0.054 | 0.694 | — | |
| STream3R-STACType=Online2026.03 | 0.198 | 0.082 | 1.486 | — | |
| TTT3RStream=true2025.12 | 0.201 | 0.063 | 0.617 | — | |
| TTT3R2026.03 | 0.201 | 0.063 | 0.671 | — | |
| TTT3Rmode=streaming2026.03 | 0.204 | 0.085 | 0.69 | — | |
| TTT3RInference Speed Complexity=O(N)2026.03 | 0.204 | 0.085 | 0.69 | — | |
| FLARE2026.02 | 0.207 | 0.09 | 3.015 | — | |
| FLARE2026.03 | 0.207 | — | 3.015 | 0.09 | |
| FLAREInference Speed Complexity=O(N^2)2026.03 | 0.207 | 0.09 | 3.015 | — | |
| FLARE2025.07 | 0.207 | 0.09 | 3.015 | — | |
| SSR2026.03 | 0.209 | 0.077 | 0.631 | — | |
| CUT3R2026.03 | 0.209 | 0.071 | 0.632 | — | |
| TTT3RType=Stream2026.01 | 0.21 | 0.09 | 0.722 | — | |
| TTSA3RType=Stream2026.01 | 0.21 | 0.085 | 0.765 | — | |
| TTT3R2026.03 | 0.21 | 0.09 | 0.73 | — | |
| MapAnythingtype=FA2026.03 | 0.2104 | 0.0919 | 2.7396 | — |