Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

DDAD

Benchmarks

Task NameDataset NameSOTA ResultTrend
Monocular Depth EstimationDDAD (test)
RMSE5.307
122
Multi-view Depth EstimationDDAD (test)
AbsRel0.092
40
3D ReconstructionDDAD
Accuracy8.015
36
Monocular Depth EstimationDDAD
Abs Rel Error0.077
33
Depth EstimationDDAD (val)
Sq Rel1.29
31
Depth EstimationDDAD
Delta1 Error0.272
26
Depth CompletionDDAD
MAE1.344
23
Sample-wise Anomaly DetectionDDAD (test)
AP84
15
Cross-view depth consistencyDDAD
Abs Rel0.162
12
Global Point Map EstimationDDAD
Absolute Relative Error (Global)9.05
12
Local Point Map PredictionDDAD
AbsRel (loc)6.29
12
Depth EstimationDDAD zero-shot
RMSE5.365
11
Monocular Depth EstimationDDAD outdoor (test)
Delta < 1.25^3 Accuracy99
10
Pixel-wise Anomaly DetectionDDAD (test)
AP24.3
9
Geometric ReconstructionDDAD (test)
Relp14.58
8
Camera Pose EstimationDDAD
AUC@30°97.4
7
Inference EfficiencyDDAD (val)
Latency (ms)58.83
7
Depth EstimationDDAD 12 (test)
Abs Rel0.088
6
Ego-pose EstimationDDAD
AUC@3095.1
5
Depth estimation in overlapping regionsDDAD
Abs Rel Error0.207
4
Relative Depth EstimationDDAD 20 (test)
AbsRel0.23
4
Single-view 3D Scene ReconstructionDDAD (test)
Oacc59
3
Depth CompletionDDAD (test)
RMSE399.2
3
Monocular Depth EstimationDDAD Dynamic
Abs Rel0.234
3
Monocular Depth EstimationDDAD (Overall)
Abs Rel0.158
3
Showing 25 of 27 rows