Hidden

FTIR Spectroscopic Imaging — Hidden Tier

(5 scenes)

Fully blind server-side evaluation — no data download.

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.

Parameter Specifications

🔒

True spec hidden — blind evaluation, only ranges available.

Parameter Spec Range Unit
wavenumber_calibration -0.28 – 0.92 cm^-1
water_vapor_absorption -0.15 – 0.15 -
detector_nonlinearity -0.7 – 2.3 -
atr_crystal_ri_error -0.14 – 0.46 -

Hidden Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 PINN-Spectra + gradient 0.650 25.95 0.831 0.79 ✓ Certified Physics-informed neural network
2 Cascade-UNet + gradient 0.579 22.41 0.707 0.86 ✓ Certified Physics-informed UNet, 2025
3 Baseline Correction + gradient 0.567 22.4 0.707 0.8 ✓ Certified Polynomial fitting baseline
4 SpectraFormer + gradient 0.558 22.37 0.706 0.76 ✓ Certified Spectroscopy transformer, 2024
5 DiffusionSpectra + gradient 0.554 22.15 0.696 0.77 ✓ Certified Zhang et al., 2024
6 CDAE + gradient 0.549 21.47 0.667 0.84 ✓ Certified Zhang et al., Sensors 2024
7 U-Net-Spectra + gradient 0.544 21.34 0.661 0.83 ✓ Certified Spectral U-Net variant
8 SVD + gradient 0.526 21.27 0.658 0.75 ✓ Certified Singular Value Decomposition
9 PnP-DnCNN + gradient 0.520 20.43 0.619 0.84 ✓ Certified Zhang et al., 2017
10 SG-ALS + gradient 0.495 20.12 0.604 0.76 ✓ Certified Savitzky-Golay + ALS baseline
11 ScoreSpectra + gradient 0.430 17.32 0.466 0.85 ✓ Certified Wei et al., 2025

Dataset

Scenes: 5

Scoring Formula

0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖)

PSNR: 40% SSIM: 40% Consistency: 20%
Back to FTIR Spectroscopic Imaging