Public
Hyperspectral Remote Sensing — 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 |
|---|---|---|---|
| spectral_shift | -0.4 – 0.8 | 0.2 | nm |
| smile_distortion | -0.2 – 0.4 | 0.1 | px |
| keystone_distortion | -0.1 – 0.2 | 0.05 | px |
| radiometric_gain | 0.98 – 1.04 | 1.01 | - |
InverseNet Baseline Scores
Method: CPU_baseline — Mismatch parameter: nominal
Scenario I (Ideal)
9.47 dB
SSIM 0.1882
Scenario II (Mismatch)
9.32 dB
SSIM 0.1839
Scenario III (Oracle)
11.90 dB
SSIM 0.2259
Per-scene breakdown (4 scenes)
| Scene | PSNR I | SSIM I | PSNR II | SSIM II | PSNR III | SSIM III |
|---|---|---|---|---|---|---|
| scene_00 | 10.43 | 0.6652 | 9.73 | 0.6398 | 11.21 | 0.6927 |
| scene_01 | 9.15 | 0.0294 | 9.16 | 0.0324 | 12.11 | 0.0709 |
| scene_02 | 9.12 | 0.0284 | 9.22 | 0.0318 | 12.17 | 0.0694 |
| scene_03 | 9.15 | 0.0298 | 9.16 | 0.0316 | 12.11 | 0.0708 |
Public Tier Leaderboard
| # | Method | Score | PSNR | SSIM | Consistency | Trust | Source |
|---|---|---|---|---|---|---|---|
| 1 | MST++ + gradient | 0.828 | 35.21 | 0.969 | 0.94 | ✓ Certified | Cai et al., CVPRW 2022 |
| 2 | DBIN + gradient | 0.774 | 31.76 | 0.94 | 0.9 | ✓ Certified | Dong et al., CVPR 2021 |
| 3 | PnP-LTTR + gradient | 0.699 | 27.17 | 0.862 | 0.91 | ✓ Certified | He et al., IEEE TGRS 2020 |
| 4 | CNMF + gradient | 0.623 | 24.06 | 0.771 | 0.87 | ✓ Certified | Yokoya et al., IEEE TGRS 2012 |
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%