Hidden

OCT — 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
dispersion -140.0 – 460.0 fs²
reference_delay -3.5 – 11.5 μm
spectral_roll_off -0.7 – 2.3 dB/mm

Hidden Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 SpeckleFormer + gradient 0.731 29.87 0.915 0.83 ✓ Certified Devalla et al., ECCV 2024
2 RetinalFormer + gradient 0.723 30.02 0.917 0.78 ✓ Certified Chen et al., ICCV 2024
3 OCT-ViT + gradient 0.723 30.29 0.921 0.76 ✓ Certified Tian et al., ICCV 2024
4 ScoreOCT + gradient 0.705 29.4 0.907 0.74 ✓ Certified Wei et al., ECCV 2025
5 DiffusionOCT + gradient 0.682 27.32 0.866 0.81 ✓ Certified Zhang et al., NeurIPS 2024
6 Speckle-Lee + gradient 0.651 25.71 0.824 0.82 ✓ Certified Lee, IEEE TGRS 1980
7 BM4D + gradient 0.612 23.51 0.751 0.88 ✓ Certified Maggioni et al., IEEE TIP 2013
8 Speckle-DenoiseNet + gradient 0.596 23.87 0.764 0.76 ✓ Certified Devalla et al., BOE 2019
9 TV-Denoising + gradient 0.595 22.86 0.726 0.88 ✓ Certified Rudin et al., Phys. A 1992
10 OCTA-Net + gradient 0.593 22.87 0.726 0.87 ✓ Certified Hybrid U-Net+Transformer, 2023
11 FFT-OCT + gradient 0.539 21.62 0.673 0.77 ✓ Certified Analytical baseline
12 NLM-OCT + gradient 0.521 20.32 0.614 0.86 ✓ Certified Buades et al., Multiscale Model. Simul. 2005
13 U-Net-OCT + gradient 0.507 20.6 0.627 0.75 ✓ Certified Ronneberger et al., MICCAI 2015 (OCT variant)

Dataset

Scenes: 3

Scoring Formula

0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖)

PSNR: 40% SSIM: 40% Consistency: 20%
Back to OCT