Hidden

TIRF — 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
incidence_angle -0.21 – 0.69 deg
penetration_depth -14.0 – 46.0 nm
refractive_index 1.5115 – 1.5265

Hidden Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 DeconvFormer + gradient 0.719 29.54 0.91 0.8 ✓ Certified Chen et al., CVPR 2024
2 Restormer+ + gradient 0.689 27.94 0.88 0.79 ✓ Certified Zamir et al., ICCV 2024
3 DiffDeconv + gradient 0.658 26.56 0.847 0.77 ✓ Certified Huang et al., NeurIPS 2024
4 ScoreMicro + gradient 0.654 25.98 0.831 0.81 ✓ Certified Wei et al., ECCV 2025
5 PnP-FISTA + gradient 0.654 25.86 0.828 0.82 ✓ Certified Bai et al., 2020
6 TV-Deconvolution + gradient 0.641 25.44 0.816 0.8 ✓ Certified Rudin et al., Phys. A 1992
7 Restormer + gradient 0.617 24.08 0.771 0.84 ✓ Certified Zamir et al., CVPR 2022
8 Wiener Filter + gradient 0.587 22.62 0.716 0.87 ✓ Certified Analytical baseline
9 CARE + gradient 0.583 22.48 0.71 0.87 ✓ Certified Weigert et al., Nat. Methods 2018
10 PnP-DnCNN + gradient 0.566 22.59 0.715 0.77 ✓ Certified Zhang et al., IEEE TIP 2017
11 ResUNet + gradient 0.563 22.05 0.692 0.83 ✓ Certified DeCelle et al., Nat. Methods 2021
12 U-Net + gradient 0.559 22.4 0.707 0.76 ✓ Certified Ronneberger et al., MICCAI 2015
13 Richardson-Lucy + gradient 0.550 21.99 0.69 0.77 ✓ Certified Richardson, JOSA 1972 / Lucy, AJ 1974

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 TIRF