Public

Electron Tomo — Public Tier

(3 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
tilt_angle -0.5 – 1.0 0.25 deg
tilt_axis -0.3 – 0.6 0.15 deg
defocus_gradient -10.0 – 20.0 5.0 nm/μm

InverseNet Baseline Scores

Method: CPU_baseline — Mismatch parameter: nominal

Scenario I (Ideal)

16.13 dB

SSIM 0.1412

Scenario II (Mismatch)

15.15 dB

SSIM 0.0535

Scenario III (Oracle)

17.70 dB

SSIM 0.1533

Per-scene breakdown (4 scenes)
Scene PSNR I SSIM I PSNR II SSIM II PSNR III SSIM III
scene_00 15.51 0.1325 14.81 0.0541 18.01 0.1778
scene_01 16.91 0.1523 15.32 0.0536 17.33 0.1392
scene_02 16.78 0.1493 15.75 0.0564 17.89 0.1431
scene_03 15.33 0.1308 14.71 0.0498 17.58 0.1532

Public Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 PhysET + gradient 0.839 36.16 0.974 0.94 ✓ Certified Chen et al., Nat. Commun. 2024
2 DiffET + gradient 0.834 36.57 0.976 0.89 ✓ Certified Gao et al., NeurIPS 2024
3 SwinET + gradient 0.803 34.36 0.963 0.87 ✓ Certified Wang et al., Ultramicroscopy 2023
4 TransET + gradient 0.802 33.4 0.956 0.93 ✓ Certified Li et al., Nat. Methods 2022
5 IsoNet + gradient 0.768 31.1 0.932 0.92 ✓ Certified Liu et al., Nat. Commun. 2021
6 DnCNN-ET + gradient 0.691 27.07 0.86 0.88 ✓ Certified Buchholz et al., Nat. Methods 2019
7 CS-ET + gradient 0.617 23.49 0.75 0.91 ✓ Certified Leary et al., Ultramicroscopy 2013
8 SIRT-ET + gradient 0.591 22.41 0.707 0.92 ✓ Certified Gilbert, J. Theor. Biol. 1972
9 WBP-ET + gradient 0.460 17.91 0.496 0.91 ✓ Certified Radermacher et al., J. Microsc. 1987

Visible Data Fields

y H_ideal spec_ranges x_true true_spec

Dataset

Format: HDF5
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 Electron Tomo