Dev

Widefield Low-Dose — Dev Tier

(5 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
psf_sigma -12.0 – 18.0 %
photon_budget -24.0 – 36.0 %
read_noise 0.3 – 3.3 e-

Dev Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 DeconvFormer + gradient 0.785 33.93 0.96 0.81 ✓ Certified Chen et al., CVPR 2024
2 ScoreMicro + gradient 0.765 31.97 0.942 0.84 ✓ Certified Wei et al., ECCV 2025
3 Restormer+ + gradient 0.755 31.56 0.938 0.82 ✓ Certified Zamir et al., ICCV 2024
4 DiffDeconv + gradient 0.753 30.65 0.926 0.88 ✓ Certified Huang et al., NeurIPS 2024
5 Restormer + gradient 0.719 29.51 0.909 0.8 ✓ Certified Zamir et al., CVPR 2022
6 ResUNet + gradient 0.715 28.46 0.89 0.87 ✓ Certified DeCelle et al., Nat. Methods 2021
7 U-Net + gradient 0.707 28.15 0.884 0.86 ✓ Certified Ronneberger et al., MICCAI 2015
8 TV-Deconvolution + gradient 0.672 26.37 0.842 0.86 ✓ Certified Rudin et al., Phys. A 1992
9 PnP-DnCNN + gradient 0.653 25.81 0.827 0.82 ✓ Certified Zhang et al., IEEE TIP 2017
10 CARE + gradient 0.652 25.3 0.812 0.87 ✓ Certified Weigert et al., Nat. Methods 2018
11 PnP-FISTA + gradient 0.638 25.41 0.815 0.79 ✓ Certified Bai et al., 2020
12 Wiener Filter + gradient 0.633 24.39 0.782 0.88 ✓ Certified Analytical baseline
13 Richardson-Lucy + gradient 0.615 23.49 0.75 0.9 ✓ Certified Richardson, JOSA 1972 / Lucy, AJ 1974

Visible Data Fields

y H_ideal spec_ranges

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%
Back to Widefield Low-Dose