Interferometric SAR (InSAR)

Interferometric SAR (InSAR)

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
🥇 InSAR-Former 0.760 33.0 0.920 ✓ Certified InSAR phase transformer, 2024
🥈 PhaseNet 0.712 31.0 0.890 ✓ Certified Sica et al., IEEE TGRS 2021
🥉 InSAR-BM3D 0.595 27.0 0.790 ✓ Certified Deledalle et al., IEEE TIP 2015
4 Goldstein-MCF 0.453 23.0 0.640 ✓ Certified Goldstein et al., Radio Sci. 1988

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
🥇 InSAR-Former + gradient 0.713
0.749
30.02 dB / 0.917
0.713
29.14 dB / 0.903
0.676
27.26 dB / 0.864
✓ Certified InSAR phase transformer, 2024
🥈 PhaseNet + gradient 0.596
0.721
28.62 dB / 0.893
0.593
23.09 dB / 0.735
0.473
18.63 dB / 0.531
✓ Certified Sica et al., IEEE TGRS 2021
🥉 InSAR-BM3D + gradient 0.539
0.645
24.9 dB / 0.799
0.523
20.59 dB / 0.627
0.448
18.27 dB / 0.513
✓ Certified Deledalle et al., IEEE TIP 2015
4 Goldstein-MCF + gradient 0.522
0.526
20.16 dB / 0.606
0.553
21.59 dB / 0.672
0.487
19.93 dB / 0.595
✓ Certified Goldstein et al., Radio Sci. 1988

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 InSAR-Former + gradient 0.749 30.02 0.917
2 PhaseNet + gradient 0.721 28.62 0.893
3 InSAR-BM3D + gradient 0.645 24.9 0.799
4 Goldstein-MCF + gradient 0.526 20.16 0.606
Spec Ranges (4 parameters)
Parameter Min Max Unit
phase_unwrapping_error -1.0 2.0 -
baseline_estimation_error -0.2 0.4 m
atmospheric_phase_screen -0.2 0.4 radrms
temporal_decorrelation -0.06 0.12 coherenceloss
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 InSAR-Former + gradient 0.713 29.14 0.903
2 PhaseNet + gradient 0.593 23.09 0.735
3 Goldstein-MCF + gradient 0.553 21.59 0.672
4 InSAR-BM3D + gradient 0.523 20.59 0.627
Spec Ranges (4 parameters)
Parameter Min Max Unit
phase_unwrapping_error -1.2 1.8 -
baseline_estimation_error -0.24 0.36 m
atmospheric_phase_screen -0.24 0.36 radrms
temporal_decorrelation -0.072 0.108 coherenceloss
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 InSAR-Former + gradient 0.676 27.26 0.864
2 Goldstein-MCF + gradient 0.487 19.93 0.595
3 PhaseNet + gradient 0.473 18.63 0.531
4 InSAR-BM3D + gradient 0.448 18.27 0.513
Spec Ranges (4 parameters)
Parameter Min Max Unit
phase_unwrapping_error -0.7 2.3 -
baseline_estimation_error -0.14 0.46 m
atmospheric_phase_screen -0.14 0.46 radrms
temporal_decorrelation -0.042 0.138 coherenceloss
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

F → S → D

F Fourier
S Sampling
D Detector

Mismatch Parameters

Symbol Parameter Description Nominal Perturbed
p_u phase_unwrapping_error Phase unwrapping error (-) 0.0 1.0
b_e baseline_estimation_error Baseline estimation error (m) 0.0 0.2
a_p atmospheric_phase_screen Atmospheric phase screen (rad rms) 0.0 0.2
t_d temporal_decorrelation Temporal decorrelation (coherence loss) 0.0 0.06

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.