Public

STED — Public Tier

(5 scenes)

Full-access development tier with all data visible.

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.

Parameter Specifications

✓

True spec visible — use these exact values for Scenario III oracle reconstruction.

Parameter Spec Range True Value Unit
depletion_power -10.0 – 20.0 5.0 %
donut_alignment -10.0 – 20.0 5.0 nm
saturation_intensity -8.0 – 16.0 4.0 %

InverseNet Baseline Scores

Method: CPU_baseline — Mismatch parameter: nominal

Scenario I (Ideal)

28.91 dB

SSIM 0.6861

Scenario II (Mismatch)

25.74 dB

SSIM 0.5319

Scenario III (Oracle)

6.98 dB

SSIM 0.0016

Per-scene breakdown (4 scenes)
Scene PSNR I SSIM I PSNR II SSIM II PSNR III SSIM III
scene_00 25.01 0.7227 21.68 0.6331 8.51 0.0030
scene_01 33.59 0.8983 28.55 0.6802 8.56 0.0010
scene_02 28.66 0.3828 27.04 0.2198 5.30 0.0013
scene_03 28.37 0.7405 25.68 0.5944 5.57 0.0012

Public Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 DiffDeconv + gradient 0.846 37.06 0.978 0.92 ✓ Certified Huang et al., NeurIPS 2024
2 ScoreMicro + gradient 0.827 35.99 0.973 0.89 ✓ Certified Wei et al., ECCV 2025
3 Restormer+ + gradient 0.816 34.8 0.966 0.91 ✓ Certified Zamir et al., ICCV 2024
4 ResUNet + gradient 0.816 34.28 0.963 0.94 ✓ Certified DeCelle et al., Nat. Methods 2021
5 DeconvFormer + gradient 0.812 34.64 0.965 0.9 ✓ Certified Chen et al., CVPR 2024
6 U-Net + gradient 0.808 33.99 0.961 0.92 ✓ Certified Ronneberger et al., MICCAI 2015
7 Restormer + gradient 0.795 33.73 0.959 0.87 ✓ Certified Zamir et al., CVPR 2022
8 CARE + gradient 0.776 31.9 0.942 0.9 ✓ Certified Weigert et al., Nat. Methods 2018
9 PnP-DnCNN + gradient 0.748 29.56 0.91 0.94 ✓ Certified Zhang et al., IEEE TIP 2017
10 TV-Deconvolution + gradient 0.718 27.77 0.876 0.95 ✓ Certified Rudin et al., Phys. A 1992
11 PnP-FISTA + gradient 0.711 28.02 0.881 0.89 ✓ Certified Bai et al., 2020
12 Wiener Filter + gradient 0.698 26.92 0.856 0.93 ✓ Certified Analytical baseline
13 Richardson-Lucy + gradient 0.673 25.66 0.822 0.94 ✓ Certified Richardson, JOSA 1972 / Lucy, AJ 1974

Visible Data Fields

y H_ideal spec_ranges x_true true_spec

Dataset

Format: HDF5
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 STED