Hidden
OCTA — Hidden Tier
(3 scenes)Fully blind server-side evaluation — no data download.
What you get
No data downloadable. Algorithm runs server-side on hidden measurements.
How to use
Package algorithm as Docker container / Python script. Submit via link.
What to submit
Containerized algorithm accepting y + H, outputting x_hat + corrected spec.
Parameter Specifications
🔒
True spec hidden — blind evaluation, only ranges available.
| Parameter | Spec Range | Unit |
|---|---|---|
| inter_bscan_time | -0.35 – 1.15 | ms |
| bulk_motion | -0.14 – 0.46 | mm/s |
| decorrelation_threshold | 0.465 – 0.615 |
Hidden Tier Leaderboard
| # | Method | Score | PSNR | SSIM | Consistency | Trust | Source |
|---|---|---|---|---|---|---|---|
| 1 | OCT-ViT + gradient | 0.739 | 30.66 | 0.926 | 0.81 | ✓ Certified | Tian et al., ICCV 2024 |
| 2 | DiffusionOCT + gradient | 0.725 | 30.14 | 0.919 | 0.78 | ✓ Certified | Zhang et al., NeurIPS 2024 |
| 3 | ScoreOCT + gradient | 0.715 | 28.35 | 0.888 | 0.88 | ✓ Certified | Wei et al., ECCV 2025 |
| 4 | SpeckleFormer + gradient | 0.652 | 25.68 | 0.823 | 0.83 | ✓ Certified | Devalla et al., ECCV 2024 |
| 5 | OCTA-Net + gradient | 0.646 | 25.97 | 0.831 | 0.77 | ✓ Certified | Hybrid U-Net+Transformer, 2023 |
| 6 | RetinalFormer + gradient | 0.629 | 24.48 | 0.785 | 0.85 | ✓ Certified | Chen et al., ICCV 2024 |
| 7 | NLM-OCT + gradient | 0.608 | 24.35 | 0.781 | 0.76 | ✓ Certified | Buades et al., Multiscale Model. Simul. 2005 |
| 8 | TV-Denoising + gradient | 0.608 | 24.52 | 0.787 | 0.74 | ✓ Certified | Rudin et al., Phys. A 1992 |
| 9 | BM4D + gradient | 0.578 | 22.81 | 0.724 | 0.8 | ✓ Certified | Maggioni et al., IEEE TIP 2013 |
| 10 | Speckle-Lee + gradient | 0.575 | 22.2 | 0.698 | 0.87 | ✓ Certified | Lee, IEEE TGRS 1980 |
| 11 | U-Net-OCT + gradient | 0.559 | 22.12 | 0.695 | 0.8 | ✓ Certified | Ronneberger et al., MICCAI 2015 (OCT variant) |
| 12 | FFT-OCT + gradient | 0.527 | 21.1 | 0.65 | 0.78 | ✓ Certified | Analytical baseline |
| 13 | Speckle-DenoiseNet + gradient | 0.500 | 20.43 | 0.619 | 0.74 | ✓ Certified | Devalla et al., BOE 2019 |
Dataset
Scenes: 3
Scoring Formula
0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖)
PSNR: 40%
SSIM: 40%
Consistency: 20%