Helen
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Face Parsing | Helen (test) | Eyes Accuracy0.909 | 22 | |
| Face Hallucination | Helen | PSNR27.95 | 16 | |
| Face Deblurring | Helen | PSNR27.75 | 11 | |
| Facial Landmark Detection | Helen 194 landmarks Original | Mean Error Rate4.63 | 9 | |
| Facial Landmark Detection | Helen 68 landmarks 300-W protocol | Mean Error (%)4.6 | 8 | |
| Face Parsing | Helen | Nose Accuracy96.3 | 6 | |
| Face Deblurring | Helen | dVGG Score12.28 | 6 | |
| Malicious Editing | Helen (test) | SSIM0.52 | 5 | |
| Facial Landmark Detection | Helen | Latency (ms)18 | 3 | |
| Portrait Drawing Generation | Helen (test) | User Preference64.2 | 1 |
Showing 10 of 10 rows