Dev

Holography — Dev Tier

(5 scenes)

Blind evaluation tier — no ground truth available.

What you get

Measurements (y), ideal forward operator (H), and spec ranges only.

How to use

Apply your pipeline from the Public tier. Use consistency as self-check.

What to submit

Reconstructed signals and corrected spec. Scored server-side.

Parameter Specifications

🔒

True spec hidden — estimate parameters from spec ranges below.

Parameter Spec Range Unit
wavelength -0.6 – 0.9 nm
prop_distance -6.0 – 9.0 μm
tilt -0.6 – 0.9 mrad

Dev Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 AutoPhase++ + gradient 0.713 28.23 0.886 0.88 ✓ Certified Rivenson et al., ECCV 2024
2 HolographyViT + gradient 0.712 28.44 0.89 0.86 ✓ Certified Wang et al., ICCV 2024
3 PhaseFormer + gradient 0.692 27.12 0.861 0.88 ✓ Certified Tian et al., ICCV 2024
4 ScorePhase + gradient 0.682 27.56 0.871 0.79 ✓ Certified Wei et al., ECCV 2025
5 CyclePhase + gradient 0.670 26.43 0.844 0.84 ✓ Certified Ge et al., IEEE Photonics 2023
6 PhaseResNet + gradient 0.655 25.73 0.824 0.84 ✓ Certified Baoqing et al., Optica 2023
7 DiffusionPhase + gradient 0.642 24.96 0.801 0.86 ✓ Certified Song et al., NeurIPS 2024
8 LRGS + gradient 0.596 22.74 0.721 0.9 ✓ Certified Choi et al., 2023
9 PhaseNet + gradient 0.587 22.62 0.716 0.87 ✓ Certified Rivenson et al., LSA 2018
10 prDeep + gradient 0.537 21.23 0.656 0.81 ✓ Certified Metzler et al., ICML 2018
11 GS/HIO + gradient 0.534 20.58 0.626 0.89 ✓ Certified Fienup, Appl. Opt. 1982
12 deep-PR + gradient 0.502 19.73 0.586 0.85 ✓ Certified Asif et al., ICCP 2017
13 Gerchberg-Saxton + gradient 0.495 19.63 0.581 0.83 ✓ Certified Gerchberg & Saxton, Optik 1972
14 Error Reduction + gradient 0.481 18.69 0.534 0.9 ✓ Certified Fienup, J. Opt. Soc. Am. 1982

Visible Data Fields

y H_ideal spec_ranges

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