Dev
4D-STEM — Dev Tier
(3 scenes)Blind evaluation tier — no ground truth available.
What you get
Measurements (y), ideal forward operator (H), and spec ranges only.
How to use
Apply your pipeline from the Public tier. Use consistency as self-check.
What to submit
Reconstructed signals and corrected spec. Scored server-side.
Parameter Specifications
🔒
True spec hidden — estimate parameters from spec ranges below.
| Parameter | Spec Range | Unit |
|---|---|---|
| camera_length | -2.4 – 3.6 | % |
| center_offset | -1.2 – 1.8 | pixels |
| elliptical_distortion | -0.006 – 0.009 |
Dev Tier Leaderboard
| # | Method | Score | PSNR | SSIM | Consistency | Trust | Source |
|---|---|---|---|---|---|---|---|
| 1 | SwinED + gradient | 0.755 | 30.61 | 0.926 | 0.89 | ✓ Certified | Wang et al., npj Comput. Mater. 2023 |
| 2 | DiffED + gradient | 0.742 | 31.09 | 0.932 | 0.79 | ✓ Certified | Gao et al., NeurIPS 2024 |
| 3 | PhysED + gradient | 0.732 | 30.43 | 0.923 | 0.79 | ✓ Certified | Chen et al., Nat. Commun. 2024 |
| 4 | TransED + gradient | 0.713 | 28.13 | 0.883 | 0.89 | ✓ Certified | Li et al., Nat. Commun. 2022 |
| 5 | PhaseGAN-ED + gradient | 0.621 | 24.33 | 0.78 | 0.83 | ✓ Certified | Zimmermann et al., Sci. Adv. 2021 |
| 6 | MicroED + gradient | 0.609 | 23.56 | 0.753 | 0.86 | ✓ Certified | Shi et al., eLife 2013 |
| 7 | DnCNN-ED + gradient | 0.522 | 20.29 | 0.613 | 0.87 | ✓ Certified | Cherukara et al., npj Comput. Mater. 2018 |
| 8 | PEDT + gradient | 0.512 | 20.43 | 0.619 | 0.8 | ✓ Certified | Kolb et al., Ultramicroscopy 2007 |
| 9 | Direct-Methods + gradient | 0.454 | 18.46 | 0.523 | 0.8 | ✓ Certified | Hauptman & Karle, Nobel Prize 1985 |
Visible Data Fields
y
H_ideal
spec_ranges
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%