Hidden

Holography — Hidden Tier

(5 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
wavelength -0.35 – 1.15 nm
prop_distance -3.5 – 11.5 μm
tilt -0.35 – 1.15 mrad

Hidden Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 ScorePhase + gradient 0.662 25.67 0.823 0.88 ✓ Certified Wei et al., ECCV 2025
2 AutoPhase++ + gradient 0.646 25.88 0.829 0.78 ✓ Certified Rivenson et al., ECCV 2024
3 HolographyViT + gradient 0.644 25.86 0.828 0.77 ✓ Certified Wang et al., ICCV 2024
4 DiffusionPhase + gradient 0.620 24.52 0.787 0.8 ✓ Certified Song et al., NeurIPS 2024
5 CyclePhase + gradient 0.599 23.11 0.735 0.87 ✓ Certified Ge et al., IEEE Photonics 2023
6 PhaseFormer + gradient 0.596 22.97 0.73 0.87 ✓ Certified Tian et al., ICCV 2024
7 PhaseResNet + gradient 0.562 21.87 0.684 0.85 ✓ Certified Baoqing et al., Optica 2023
8 PhaseNet + gradient 0.515 20.96 0.644 0.74 ✓ Certified Rivenson et al., LSA 2018
9 LRGS + gradient 0.478 19.68 0.583 0.74 ✓ Certified Choi et al., 2023
10 GS/HIO + gradient 0.474 18.87 0.543 0.84 ✓ Certified Fienup, Appl. Opt. 1982
11 prDeep + gradient 0.459 18.23 0.511 0.86 ✓ Certified Metzler et al., ICML 2018
12 Gerchberg-Saxton + gradient 0.448 17.85 0.493 0.86 ✓ Certified Gerchberg & Saxton, Optik 1972
13 Error Reduction + gradient 0.404 16.59 0.43 0.83 ✓ Certified Fienup, J. Opt. Soc. Am. 1982
14 deep-PR + gradient 0.401 16.69 0.435 0.8 ✓ Certified Asif et al., ICCP 2017

Dataset

Scenes: 5

Scoring Formula

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

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