| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Semantic Segmentation | Structured3D (test) | mIoU82.4 | 21 | |
| Semantic Segmentation | Structured3D (val) | mIoU85.4 | 17 | |
| Floorplan Localization | Structured3D (full) | Recall @ 0.1m61 | 15 | |
| Floorplan Reconstruction | Structured3D density map input (test) | Room Precision99.6 | 11 | |
| Scene Layout Estimation | Structured3D (test) | F1 Score (wall)80.3 | 10 | |
| Floorplan Reconstruction | Structured3D binary (test) | Room F199.6 | 10 | |
| Depth Estimation | Structured3D (val) | δ1 Accuracy96.79 | 9 | |
| 6D camera localization | Structured3D Furnishing-Level: Full | Median Translation Error (<1m)9.17 | 9 | |
| View Synthesis | Structured3D Hard Set (1.0 m to 2.0 m) 1.0 | PSNR18.95 | 9 | |
| View Synthesis | Structured3D Easy Set (0.2 m to 0.3 m) 1.0 | PSNR20.83 | 9 | |
| Surface Normal Estimation | Structured3D (test) | Mean Angular Error (deg)3.85 | 8 | |
| Perspective 120° FoV image-to-map localization | Structured3D Furnishing-Level: Full | Recall @ 10cm30.88 | 6 | |
| Panorama image-to-map localization | Structured3D Furnishing-Level: Full | Median Terr (<1m) [cm]3.87 | 6 | |
| Monocular Depth Estimation | Structured3D | MAE0.0454 | 6 | |
| Semantic Segmentation | Structured3D sphere rank 7 256x512 (test) | Accuracy95.8 | 5 | |
| Depth Estimation | Structured3D sphere rank 7 256x512 (test) | MAE0.142 | 5 | |
| Floorplan Localization | Structured3D 69 | Acc (0.1m, 5°)95 | 5 | |
| Perspective 90° FoV image-to-map localization | Structured3D Furnishing-Level Full | Median Translational Error (1m) [cm]12.99 | 4 | |
| Perspective 60° FoV image-to-map localization | Structured3D Furnishing-Level: Full | Median terr (<1m) [cm]16.97 | 4 | |
| Monocular Panoramic Depth Estimation | Structured3D (official) | MRE (Mean Relative Error)0.0376 | 4 | |
| Semantic Segmentation | Structured3D S3D8 | mIoU0.7729 | 3 | |
| Floorplan Semantic Segmentation | Structured3D 50 (test) | Room Semantic Prec.76.8 | 2 |