300-W
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Facial Landmark Detection | 300-W (Common) | NME0.0241 | 180 | |
| Facial Landmark Detection | 300-W (Fullset) | Mean Error (%)2.78 | 174 | |
| Face Alignment | 300-W (Full) | NME2.87 | 66 | |
| Facial Landmark Detection | 300-W Challenging Subset | Mean Error4.93 | 49 | |
| Facial Landmark Localization | 300-W (Full set) | NME2.83 | 46 | |
| Facial Landmark Detection | 300-W public Challenging inter-pupil normalization (test) | NME6.38 | 46 | |
| Facial Landmark Detection | 300-W Common Subset | Mean Error (%)2.46 | 37 | |
| Facial Landmark Localization | 300-W (Challenging set) | NME3.36 | 32 | |
| Facial Landmark Detection | 300-W public Full inter-pupil normalization (test) | NME4.04 | 29 | |
| Facial Landmark Detection | 300-W Public Common subset inter-pupil normalization (test) | NME3.27 | 28 | |
| Facial Landmark Detection | 300-W public (test) | NME2.62 | 28 | |
| Landmark prediction | 300-W (test) | Landmark Prediction Error4.94 | 12 | |
| Face Alignment | 300-W (Split 1) | NME Common2.72 | 11 | |
| Facial Landmark Detection | 300-W competition 51-points (test) | AUC (0.08 Threshold)58.11 | 10 | |
| Face Alignment | 300-W Split 2 (test) | NME (Box)2.1 | 7 | |
| Facial Landmark Detection | 300-W competition 68-points (test) | AUC (Threshold 0.08)46.56 | 7 |
Showing 16 of 16 rows