Lucky Imaging

Lucky Imaging

Join the Competition Contribute to Benchmark

Standard reconstruction benchmark — forward model perfectly known, no calibration needed. Score = 0.5 × clip((PSNR−15)/30, 0, 1) + 0.5 × SSIM

# Method Score PSNR (dB) SSIM Source
🥇 SpeckleNet 0.737 31.22 0.934 ✓ Certified Xin et al., ApJ 2022
🥈 Drizzle 0.561 24.8 0.796 ✓ Certified Fruchter & Hook, PASP 2002
🥉 BDI 0.555 24.62 0.790 ✓ Certified Law et al., ApJ 2006
4 Shift-and-Add 0.486 22.65 0.717 ✓ Certified Fried, JOSA 1966

Dataset: PWM Benchmark (4 algorithms)

Blind Reconstruction Challenge — forward model has unknown mismatch, must calibrate from data. Score = 0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖)

# Method Overall Score Public
PSNR / SSIM
 Dev
PSNR / SSIM
 Hidden
PSNR / SSIM
 Trust Source
🥇 SpeckleNet + gradient 0.573
0.728
29.12 dB / 0.902
0.528
20.35 dB / 0.615
0.462
18.68 dB / 0.534
✓ Certified Xin et al., ApJ 2022
🥈 Drizzle + gradient 0.544
0.618
23.29 dB / 0.742
0.535
20.53 dB / 0.624
0.480
19.61 dB / 0.580
✓ Certified Fruchter & Hook, PASP 2002
🥉 Shift-and-Add + gradient 0.515
0.531
20.62 dB / 0.628
0.525
20.73 dB / 0.633
0.488
19.83 dB / 0.590
✓ Certified Fried, JOSA 1966
4 BDI + gradient 0.496
0.575
21.96 dB / 0.688
0.475
19.29 dB / 0.564
0.439
18.11 dB / 0.505
✓ Certified Law et al., ApJ 2006

Complete score requires all 3 tiers (Public + Dev + Hidden).

Join the competition →
Scoring: 0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖) PSNR 40% · SSIM 40% · Consistency 20%
Public 3 scenes

Full-access development tier with all data visible.

What you get & how to use

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.

Public Leaderboard
# Method Score PSNR SSIM
1 SpeckleNet + gradient 0.728 29.12 0.902
2 Drizzle + gradient 0.618 23.29 0.742
3 BDI + gradient 0.575 21.96 0.688
4 Shift-and-Add + gradient 0.531 20.62 0.628
Spec Ranges (4 parameters)
Parameter Min Max Unit
fried_parameter_(r0) 13.0 19.0 cm
frame_selection_threshold 2.0 26.0 -
isoplanatic_angle 4.0 7.0 arcsec
registration_error -0.1 0.2 px
View Details →
Dev 3 scenes

Blind evaluation tier — no ground truth available.

What you get & how to use

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.

Dev Leaderboard
# Method Score PSNR SSIM
1 Drizzle + gradient 0.535 20.53 0.624
2 SpeckleNet + gradient 0.528 20.35 0.615
3 Shift-and-Add + gradient 0.525 20.73 0.633
4 BDI + gradient 0.475 19.29 0.564
Spec Ranges (4 parameters)
Parameter Min Max Unit
fried_parameter_(r0) 12.6 18.6 cm
frame_selection_threshold 0.4 24.4 -
isoplanatic_angle 3.8 6.8 arcsec
registration_error -0.12 0.18 px
Submit Reconstruction View Details →
Hidden 3 scenes

Fully blind server-side evaluation — no data download.

What you get & how to use

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.

Hidden Leaderboard
# Method Score PSNR SSIM
1 Shift-and-Add + gradient 0.488 19.83 0.59
2 Drizzle + gradient 0.480 19.61 0.58
3 SpeckleNet + gradient 0.462 18.68 0.534
4 BDI + gradient 0.439 18.11 0.505
Spec Ranges (4 parameters)
Parameter Min Max Unit
fried_parameter_(r0) 13.6 19.6 cm
frame_selection_threshold 4.4 28.4 -
isoplanatic_angle 4.3 7.3 arcsec
registration_error -0.07 0.23 px
Submit Algorithm View Details →

Blind Reconstruction Challenge

Challenge

Given measurements with unknown mismatch and spec ranges (not exact params), reconstruct the original signal. A method must be evaluated on all three tiers for a complete score. Scored on a composite metric: 0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖).

Input

Measurements y, ideal forward model H, spec ranges

Output

Reconstructed signal x̂

Spec DAG — Forward Model Pipeline

M → C → D

M Modulation
C Convolution
D Detector

Mismatch Parameters

Symbol Parameter Description Nominal Perturbed
f_p fried_parameter_(r0) Fried parameter (r0) (cm) 15.0 17.0
f_s frame_selection_threshold Frame selection threshold (-) 10.0 18.0
i_a isoplanatic_angle Isoplanatic angle (arcsec) 5.0 6.0
r_e registration_error Registration error (px) 0.0 0.1

Credits System

40%
Platform Profit Pool
Revenue allocated to benchmark rewards
30%
Winner Share
Top algorithm receives from pool
$100
Min Withdrawal
Minimum payout threshold
Spec Primitives Reference (11 primitives)
P Propagation

Free-space or medium propagation kernel (Fresnel, Rayleigh-Sommerfeld).

M Mask / Modulation

Spatial or spatio-temporal amplitude modulation (coded aperture, SLM pattern).

Π Projection

Geometric projection operator (Radon transform, fan-beam, cone-beam).

F Fourier Sampling

Sampling in the Fourier / k-space domain (MRI, ptychography).

C Convolution

Shift-invariant convolution with a point-spread function (PSF).

Σ Summation / Integration

Summation along a physical dimension (spectral, temporal, angular).

D Detector

Sensor readout with gain g and noise model η (Gaussian, Poisson, mixed).

S Structured Illumination

Patterned illumination (block, Hadamard, random) applied to the scene.

W Wavelength Dispersion

Spectral dispersion element (prism, grating) with shift α and aperture a.

R Rotation / Motion

Sample or gantry rotation (CT, electron tomography).

Λ Wavelength Selection

Spectral filter or monochromator selecting a wavelength band.