Public

Fundus — Public Tier

(3 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
pupil_dilation -0.5 – 1.0 0.25 mm
focus -0.25 – 0.5 0.125 diopters
vignetting -5.0 – 10.0 2.5 %

InverseNet Baseline Scores

Method: CPU_baseline — Mismatch parameter: nominal

Scenario I (Ideal)

21.12 dB

SSIM 0.3831

Scenario II (Mismatch)

10.76 dB

SSIM 0.3611

Scenario III (Oracle)

10.61 dB

SSIM 0.0307

Per-scene breakdown (4 scenes)
Scene PSNR I SSIM I PSNR II SSIM II PSNR III SSIM III
scene_00 22.60 0.4823 10.78 0.3639 10.25 0.0292
scene_01 18.55 0.3352 11.03 0.3340 10.42 0.0310
scene_02 20.14 0.2318 10.27 0.3891 11.39 0.0358
scene_03 23.20 0.4831 10.96 0.3575 10.39 0.0265

Public Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 Swin-Fundus + gradient 0.795 32.67 0.95 0.94 ✓ Certified Li et al., IEEE TMI 2023
2 cofe-Net + gradient 0.747 30.14 0.919 0.89 ✓ Certified Shen et al., IEEE TMI 2020
3 PnP-BM3D + gradient 0.678 26.35 0.842 0.89 ✓ Certified Danielyan et al., 2012
4 Richardson-Lucy + gradient 0.611 23.1 0.735 0.93 ✓ Certified Richardson 1972 / Lucy 1974

Visible Data Fields

y H_ideal spec_ranges x_true true_spec

Dataset

Format: HDF5
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 Fundus