Public
Scanning Tunneling Microscopy (STM) — 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 |
|---|---|---|---|
| tip_electronic_structure | -0.15 – 0.15 | 0.0 | - |
| piezo_creep | -1.0 – 2.0 | 0.5 | - |
| tunneling_barrier_height | 4.2 – 5.1 | 4.65 | eV |
| vibration_amplitude | -1.0 – 2.0 | 0.5 | pm |
InverseNet Baseline Scores
Method: CPU_baseline — Mismatch parameter: nominal
Scenario I (Ideal)
9.82 dB
SSIM 0.5129
Scenario II (Mismatch)
8.93 dB
SSIM 0.1796
Scenario III (Oracle)
18.82 dB
SSIM 0.1493
Per-scene breakdown (4 scenes)
| Scene | PSNR I | SSIM I | PSNR II | SSIM II | PSNR III | SSIM III |
|---|---|---|---|---|---|---|
| scene_00 | 9.37 | 0.5121 | 9.18 | 0.1725 | 18.71 | 0.1455 |
| scene_01 | 10.22 | 0.5099 | 8.31 | 0.1922 | 18.93 | 0.1489 |
| scene_02 | 10.19 | 0.5135 | 8.93 | 0.1841 | 18.88 | 0.1581 |
| scene_03 | 9.49 | 0.5161 | 9.31 | 0.1697 | 18.75 | 0.1448 |
Public Tier Leaderboard
| # | Method | Score | PSNR | SSIM | Consistency | Trust | Source |
|---|---|---|---|---|---|---|---|
| 1 | SPM-Former + gradient | 0.782 | 32.07 | 0.943 | 0.92 | ✓ Certified | Chen et al., NanoLett 2024 |
| 2 | ScoreSPM + gradient | 0.750 | 29.82 | 0.914 | 0.93 | ✓ Certified | Wei et al., 2025 |
| 3 | E2E-BTR + gradient | 0.738 | 29.7 | 0.912 | 0.88 | ✓ Certified | Kossler et al., Sci. Rep. 2022 |
| 4 | DeepSPM + gradient | 0.711 | 28.04 | 0.882 | 0.89 | ✓ Certified | Alldritt et al., Commun. Phys. 2020 |
| 5 | DiffusionSPM + gradient | 0.711 | 27.4 | 0.868 | 0.95 | ✓ Certified | Zhang et al., 2024 |
| 6 | U-Net-SPM + gradient | 0.699 | 27.09 | 0.86 | 0.92 | ✓ Certified | SPM U-Net variant |
| 7 | TV-Deconvolution + gradient | 0.662 | 25.6 | 0.821 | 0.89 | ✓ Certified | TV regularization for SPM |
| 8 | Reg-Deconv + gradient | 0.637 | 24.58 | 0.789 | 0.88 | ✓ Certified | Dongmo et al., 2000 |
| 9 | BTR + gradient | 0.575 | 21.65 | 0.675 | 0.94 | ✓ Certified | Villarrubia, JRNIST 1997 |
| 10 | MLE Reconstruction + gradient | 0.558 | 21.4 | 0.664 | 0.89 | ✓ Certified | Classical statistical method |
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%