Ground-Penetrating Radar (GPR)

Ground-Penetrating Radar (GPR)

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
🥇 HyperDet 0.728 31.5 0.905 ✓ Certified GPR detection transformer, 2023
🥈 GPR-RCNN 0.682 29.8 0.870 ✓ Certified Pham & Lefevre, JECE 2020
🥉 RTM 0.545 25.5 0.740 ✓ Certified Baysal et al., Geophysics 1983
4 Kirchhoff Migration 0.417 22.0 0.600 ✓ Certified Stolt, Geophysics 1978

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
🥇 HyperDet + gradient 0.585
0.729
28.95 dB / 0.899
0.565
22.09 dB / 0.694
0.460
18.15 dB / 0.507
✓ Certified GPR detection transformer, 2023
🥈 RTM + gradient 0.563
0.608
23.35 dB / 0.745
0.556
22.09 dB / 0.694
0.525
21.1 dB / 0.650
✓ Certified Baysal et al., Geophysics 1983
🥉 GPR-RCNN + gradient 0.529
0.695
27.01 dB / 0.858
0.507
19.92 dB / 0.595
0.384
15.85 dB / 0.394
✓ Certified Pham & Lefevre, JECE 2020
4 Kirchhoff Migration + gradient 0.485
0.494
19.06 dB / 0.553
0.505
19.86 dB / 0.592
0.457
18.49 dB / 0.524
✓ Certified Stolt, Geophysics 1978

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 HyperDet + gradient 0.729 28.95 0.899
2 GPR-RCNN + gradient 0.695 27.01 0.858
3 RTM + gradient 0.608 23.35 0.745
4 Kirchhoff Migration + gradient 0.494 19.06 0.553
Spec Ranges (4 parameters)
Parameter Min Max Unit
soil_permittivity_error 7.8 11.4 -
antenna_height -0.01 0.02 m
time_zero_offset -0.1 0.2 ns
velocity_model_error 0.094 0.112 m/ns
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 HyperDet + gradient 0.565 22.09 0.694
2 RTM + gradient 0.556 22.09 0.694
3 GPR-RCNN + gradient 0.507 19.92 0.595
4 Kirchhoff Migration + gradient 0.505 19.86 0.592
Spec Ranges (4 parameters)
Parameter Min Max Unit
soil_permittivity_error 7.56 11.16 -
antenna_height -0.012 0.018 m
time_zero_offset -0.12 0.18 ns
velocity_model_error 0.0928 0.1108 m/ns
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 RTM + gradient 0.525 21.1 0.65
2 HyperDet + gradient 0.460 18.15 0.507
3 Kirchhoff Migration + gradient 0.457 18.49 0.524
4 GPR-RCNN + gradient 0.384 15.85 0.394
Spec Ranges (4 parameters)
Parameter Min Max Unit
soil_permittivity_error 8.16 11.76 -
antenna_height -0.007 0.023 m
time_zero_offset -0.07 0.23 ns
velocity_model_error 0.0958 0.1138 m/ns
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

P → D

P Propagation
D Detector

Mismatch Parameters

Symbol Parameter Description Nominal Perturbed
s_p soil_permittivity_error Soil permittivity error (-) 9.0 10.2
a_h antenna_height Antenna height (m) 0.0 0.01
t_z time_zero_offset Time zero offset (ns) 0.0 0.1
v_m velocity_model_error Velocity model error (m/ns) 0.1 0.106

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.