Public
Integral — 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 |
|---|---|---|---|
| lens_pitch | -1.0 – 2.0 | 0.5 | μm |
| gap_distance | -5.0 – 10.0 | 2.5 | μm |
| aberration | -0.1 – 0.2 | 0.05 | waves |
InverseNet Baseline Scores
Method: CPU_baseline — Mismatch parameter: nominal
Scenario I (Ideal)
7.39 dB
SSIM 0.3938
Scenario II (Mismatch)
7.23 dB
SSIM 0.1984
Scenario III (Oracle)
16.96 dB
SSIM 0.3543
Per-scene breakdown (4 scenes)
| Scene | PSNR I | SSIM I | PSNR II | SSIM II | PSNR III | SSIM III |
|---|---|---|---|---|---|---|
| scene_00 | 6.83 | 0.3925 | 6.75 | 0.1962 | 17.07 | 0.3460 |
| scene_01 | 7.21 | 0.3939 | 7.40 | 0.1987 | 16.92 | 0.3534 |
| scene_02 | 7.68 | 0.3938 | 7.11 | 0.2010 | 16.88 | 0.3760 |
| scene_03 | 7.84 | 0.3951 | 7.67 | 0.1976 | 16.96 | 0.3418 |
Public Tier Leaderboard
| # | Method | Score | PSNR | SSIM | Consistency | Trust | Source |
|---|---|---|---|---|---|---|---|
| 1 | DistgSSR + gradient | 0.816 | 34.13 | 0.962 | 0.95 | ✓ Certified | Wang et al., CVPR 2022 |
| 2 | LFAttNet + gradient | 0.764 | 31.65 | 0.939 | 0.86 | ✓ Certified | Tsai et al., IEEE TIP 2020 |
| 3 | PnP-LF + gradient | 0.713 | 27.73 | 0.875 | 0.93 | ✓ Certified | PnP-ADMM with LF prior |
| 4 | Shift-and-Add + gradient | 0.584 | 22.32 | 0.703 | 0.9 | ✓ Certified | Ng et al., Stanford Tech Report 2005 |
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%