Public

Holography — Public Tier

(5 scenes)

Full-access development tier with all data visible.

What you get

Measurements (y), ideal forward operator (H), spec ranges, ground truth (x_true), and true mismatch spec.

How to use

Load HDF5 → compare reconstruction vs x_true → check consistency → iterate.

What to submit

Reconstructed signals (x_hat) and corrected spec as HDF5.

Parameter Specifications

✓

True spec visible — use these exact values for Scenario III oracle reconstruction.

Parameter Spec Range True Value Unit
wavelength -0.5 – 1.0 0.25 nm
prop_distance -5.0 – 10.0 2.5 μm
tilt -0.5 – 1.0 0.25 mrad

InverseNet Baseline Scores

Method: CPU_baseline — Mismatch parameter: nominal

Scenario I (Ideal)

6.48 dB

SSIM 0.2992

Scenario II (Mismatch)

6.07 dB

SSIM 0.1771

Scenario III (Oracle)

5.02 dB

SSIM 0.0984

Per-scene breakdown (4 scenes)
Scene PSNR I SSIM I PSNR II SSIM II PSNR III SSIM III
scene_00 4.96 0.4075 4.83 0.2182 5.36 0.1110
scene_01 11.85 0.3168 10.00 0.1734 4.80 0.0434
scene_02 4.98 0.0939 4.32 0.1092 4.85 0.1279
scene_03 4.12 0.3788 5.12 0.2078 5.07 0.1111

Public Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 DiffusionPhase + gradient 0.812 34.16 0.962 0.93 ✓ Certified Song et al., NeurIPS 2024
2 ScorePhase + gradient 0.794 33.52 0.957 0.88 ✓ Certified Wei et al., ECCV 2025
3 HolographyViT + gradient 0.785 32.7 0.95 0.89 ✓ Certified Wang et al., ICCV 2024
4 AutoPhase++ + gradient 0.784 33.06 0.953 0.86 ✓ Certified Rivenson et al., ECCV 2024
5 PhaseResNet + gradient 0.780 31.48 0.937 0.95 ✓ Certified Baoqing et al., Optica 2023
6 LRGS + gradient 0.778 31.78 0.94 0.92 ✓ Certified Choi et al., 2023
7 PhaseFormer + gradient 0.777 32.29 0.946 0.88 ✓ Certified Tian et al., ICCV 2024
8 CyclePhase + gradient 0.743 29.65 0.911 0.91 ✓ Certified Ge et al., IEEE Photonics 2023
9 PhaseNet + gradient 0.728 29.1 0.902 0.88 ✓ Certified Rivenson et al., LSA 2018
10 prDeep + gradient 0.653 25.27 0.811 0.88 ✓ Certified Metzler et al., ICML 2018
11 deep-PR + gradient 0.652 25.43 0.815 0.86 ✓ Certified Asif et al., ICCP 2017
12 Error Reduction + gradient 0.574 21.76 0.68 0.92 ✓ Certified Fienup, J. Opt. Soc. Am. 1982
13 GS/HIO + gradient 0.546 20.84 0.638 0.91 ✓ Certified Fienup, Appl. Opt. 1982
14 Gerchberg-Saxton + gradient 0.488 18.94 0.547 0.9 ✓ Certified Gerchberg & Saxton, Optik 1972

Visible Data Fields

y H_ideal spec_ranges x_true true_spec

Dataset

Format: HDF5
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