Public

Lucky Imaging — 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
fried_parameter_(r0) 13.0 – 19.0 16.0 cm
frame_selection_threshold 2.0 – 26.0 14.0 -
isoplanatic_angle 4.0 – 7.0 5.5 arcsec
registration_error -0.1 – 0.2 0.05 px

InverseNet Baseline Scores

Method: CPU_baseline — Mismatch parameter: nominal

Scenario I (Ideal)

23.32 dB

SSIM 0.4205

Scenario II (Mismatch)

19.59 dB

SSIM 0.0995

Scenario III (Oracle)

21.80 dB

SSIM 0.1807

Per-scene breakdown (4 scenes)
Scene PSNR I SSIM I PSNR II SSIM II PSNR III SSIM III
scene_00 24.19 0.4390 19.09 0.0924 20.93 0.1660
scene_01 20.48 0.3616 20.39 0.1114 22.37 0.1887
scene_02 24.14 0.4400 18.87 0.0870 21.64 0.1751
scene_03 24.46 0.4415 20.01 0.1070 22.27 0.1928

Public Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 SpeckleNet + gradient 0.728 29.12 0.902 0.88 ✓ Certified Xin et al., ApJ 2022
2 Drizzle + gradient 0.618 23.29 0.742 0.94 ✓ Certified Fruchter & Hook, PASP 2002
3 BDI + gradient 0.575 21.96 0.688 0.9 ✓ Certified Law et al., ApJ 2006
4 Shift-and-Add + gradient 0.531 20.62 0.628 0.87 ✓ Certified Fried, JOSA 1966

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 Lucky Imaging