Neutron Diffraction

Neutron Diffraction

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
🥇 DiffFormer 0.749 32.5 0.915 ✓ Certified Diffraction transformer, 2024
🥈 NeutronNet 0.698 30.5 0.880 ✓ Certified Neutron diffraction DL, 2023
🥉 Le Bail Fit 0.572 26.5 0.760 ✓ Certified Le Bail et al., 1988
4 Rietveld-GSAS 0.453 23.0 0.640 ✓ Certified Rietveld, 1969

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
🥇 DiffFormer + gradient 0.691
0.747
30.12 dB / 0.919
0.706
27.66 dB / 0.873
0.621
24.99 dB / 0.802
✓ Certified Diffraction pattern transformer, 2024
🥈 NeutronNet + gradient 0.633
0.735
28.74 dB / 0.895
0.590
23.47 dB / 0.749
0.575
22.12 dB / 0.695
✓ Certified Neutron diffraction DL, 2023
🥉 Le Bail Fit + gradient 0.618
0.633
24.39 dB / 0.782
0.642
25.03 dB / 0.803
0.580
23.31 dB / 0.743
✓ Certified Le Bail et al., Mater. Res. Bull. 1988
4 Rietveld-GSAS + gradient 0.514
0.575
21.74 dB / 0.679
0.524
20.52 dB / 0.623
0.443
18.45 dB / 0.522
✓ Certified Rietveld, J. Appl. Cryst. 1969

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 5 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 DiffFormer + gradient 0.747 30.12 0.919
2 NeutronNet + gradient 0.735 28.74 0.895
3 Le Bail Fit + gradient 0.633 24.39 0.782
4 Rietveld-GSAS + gradient 0.575 21.74 0.679
Spec Ranges (4 parameters)
Parameter Min Max Unit
wavelength_calibration -0.02 0.04 -
absorption_correction -2.0 4.0 -
texture/preferred_orientation -4.0 8.0 -
tof_frame_overlap -1.0 2.0 -
View Details →
Dev 5 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 DiffFormer + gradient 0.706 27.66 0.873
2 Le Bail Fit + gradient 0.642 25.03 0.803
3 NeutronNet + gradient 0.590 23.47 0.749
4 Rietveld-GSAS + gradient 0.524 20.52 0.623
Spec Ranges (4 parameters)
Parameter Min Max Unit
wavelength_calibration -0.024 0.036 -
absorption_correction -2.4 3.6 -
texture/preferred_orientation -4.8 7.2 -
tof_frame_overlap -1.2 1.8 -
Submit Reconstruction View Details →
Hidden 5 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 DiffFormer + gradient 0.621 24.99 0.802
2 Le Bail Fit + gradient 0.580 23.31 0.743
3 NeutronNet + gradient 0.575 22.12 0.695
4 Rietveld-GSAS + gradient 0.443 18.45 0.522
Spec Ranges (4 parameters)
Parameter Min Max Unit
wavelength_calibration -0.014 0.046 -
absorption_correction -1.4 4.6 -
texture/preferred_orientation -2.8 9.2 -
tof_frame_overlap -0.7 2.3 -
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

R → S → D

R Rotation
S Sampling
D Detector

Mismatch Parameters

Symbol Parameter Description Nominal Perturbed
w_c wavelength_calibration Wavelength calibration (-) 0.0 0.02
a_c absorption_correction Absorption correction (-) 0.0 2.0
t_o texture/preferred_orientation Texture/preferred orientation (-) 0.0 4.0
t_f tof_frame_overlap TOF frame overlap (-) 0.0 1.0

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.