Dev

Widefield — Dev Tier

(5 scenes)

Blind evaluation tier — no ground truth available.

Measurements (y), ideal forward operator (H), and spec ranges only.

Apply your pipeline from the Public tier. Use consistency as self-check.

Reconstructed signals and corrected spec. Scored server-side.

🔒

True spec hidden — estimate parameters from spec ranges below.

#	Method	Score	PSNR	SSIM	Consistency	Trust	Source
1	Restormer + gradient	0.762	31.22	0.934	0.88	✓ Certified	Zamir et al., CVPR 2022
2	Restormer+ + gradient	0.762	31.9	0.942	0.83	✓ Certified	Zamir et al., ICCV 2024
3	DiffDeconv + gradient	0.757	31.85	0.941	0.81	✓ Certified	Huang et al., NeurIPS 2024
4	DeconvFormer + gradient	0.732	30.6	0.926	0.78	✓ Certified	Chen et al., CVPR 2024
5	ScoreMicro + gradient	0.726	29.8	0.914	0.81	✓ Certified	Wei et al., ECCV 2025
6	ResUNet + gradient	0.710	28.09	0.883	0.88	✓ Certified	DeCelle et al., Nat. Methods 2021
7	Wiener Filter + gradient	0.676	26.65	0.849	0.85	✓ Certified	Analytical baseline
8	U-Net + gradient	0.666	26.75	0.852	0.79	✓ Certified	Ronneberger et al., MICCAI 2015
9	TV-Deconvolution + gradient	0.649	25.27	0.811	0.86	✓ Certified	Rudin et al., Phys. A 1992
10	PnP-DnCNN + gradient	0.639	25.25	0.81	0.81	✓ Certified	Zhang et al., IEEE TIP 2017
11	PnP-FISTA + gradient	0.630	24.35	0.781	0.87	✓ Certified	Bai et al., 2020
12	Richardson-Lucy + gradient	0.608	23.94	0.766	0.81	✓ Certified	Richardson, JOSA 1972 / Lucy, AJ 1974
13	CARE + gradient	0.591	23.5	0.75	0.78	✓ Certified	Weigert et al., Nat. Methods 2018

y H_ideal spec_ranges

Format: HDF5

Scenes: 5

0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖)

PSNR: 40% SSIM: 40% Consistency: 20%