Hidden

4D-STEM — Hidden Tier

(3 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
camera_length -1.4 – 4.6 %
center_offset -0.7 – 2.3 pixels
elliptical_distortion -0.0035 – 0.0115

Hidden Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 DiffED + gradient 0.723 29.78 0.913 0.8 ✓ Certified Gao et al., NeurIPS 2024
2 PhysED + gradient 0.684 27.13 0.861 0.84 ✓ Certified Chen et al., Nat. Commun. 2024
3 SwinED + gradient 0.684 27.17 0.862 0.84 ✓ Certified Wang et al., npj Comput. Mater. 2023
4 TransED + gradient 0.612 23.77 0.76 0.85 ✓ Certified Li et al., Nat. Commun. 2022
5 MicroED + gradient 0.602 24.04 0.77 0.77 ✓ Certified Shi et al., eLife 2013
6 PhaseGAN-ED + gradient 0.536 20.8 0.636 0.87 ✓ Certified Zimmermann et al., Sci. Adv. 2021
7 PEDT + gradient 0.456 18.92 0.546 0.74 ✓ Certified Kolb et al., Ultramicroscopy 2007
8 Direct-Methods + gradient 0.437 17.37 0.469 0.88 ✓ Certified Hauptman & Karle, Nobel Prize 1985
9 DnCNN-ED + gradient 0.430 17.91 0.496 0.76 ✓ Certified Cherukara et al., npj Comput. Mater. 2018

Dataset

Scenes: 3

Scoring Formula

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

PSNR: 40% SSIM: 40% Consistency: 20%
Back to 4D-STEM