Multi-layer depth estimation on Multi-layered synthetic benchmark Transmissive Layer 1
5.47Bad-2DepthFocus
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DepthFocusInference intent=nearest-intended2025.11 | 5.47 | 3.1 | |
| DepthFocusNumber of layers=4, Aggregation strategy=clustering2025.11 | 6.24 | 3.27 | |
| S²M²Number of layers=1, Evaluation protocol=fine-tuned2025.11 | 10.9 | 7.45 | |
| S²M²Number of layers=12025.11 | 19.86 | 14.23 | |
| ASGraspNumber of layers=2, Evaluation protocol=fine-tuned2025.11 | 45.73 | 36.86 | |
| FoundationStereoNumber of layers=12025.11 | 48.97 | 40.04 | |
| ASGraspNumber of layers=4, Evaluation protocol=fine-tuned2025.11 | 57.65 | 46.49 | |
| StereoAnywhereNumber of layers=12025.11 | 62.47 | 49.68 | |
| RAFTNumber of layers=42025.11 | 70.28 | 57.09 | |
| Sel-IGEVNumber of layers=12025.11 | 72.75 | 63.02 |