Depth Estimation on Long Range Dataset (train val)
2.27MAE (0-80m)UniLiPs
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| UniLiPsPseudo source=Proposed2026.01 | 2.27 | 6.14 | 21.07 | 4.21 | 9.81 | 25.16 | |
| OraclePseudo source=naively sparse LiDAR2026.01 | 5.44 | 20.79 | 31.83 | 7.98 | 25.7 | 38.96 | |
| LIO-SAMPseudo source=LiDAR-based dense pseudo depth2026.01 | 5.53 | 11.89 | 32.33 | 10.51 | 20.11 | 46.22 | |
| DA-V2Pseudo source=Monocular foundation model2026.01 | 8.31 | 23.55 | 32.67 | 11.83 | 28.75 | 39.96 | |
| CREStereoPseudo source=Stereo prediction network2026.01 | 8.33 | 22.34 | 37.04 | 10.9 | 27.13 | 45.86 |