Public

Two-Photon — 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
pulse_width 60.0 – 180.0 120.0 fs
gdd -500.0 – 1000.0 250.0 fs²
scattering -10.0 – 20.0 5.0 %

InverseNet Baseline Scores

Method: CPU_baseline — Mismatch parameter: nominal

Scenario I (Ideal)

27.15 dB

SSIM 0.8473

Scenario II (Mismatch)

20.48 dB

SSIM 0.5701

Scenario III (Oracle)

0.00 dB

SSIM 0.0000

Per-scene breakdown (4 scenes)
Scene PSNR I SSIM I PSNR II SSIM II PSNR III SSIM III
scene_00 33.79 0.9756 20.51 0.6715 0.00 0.0000
scene_01 25.83 0.9369 25.36 0.7946 0.00 0.0000
scene_02 19.24 0.6770 18.36 0.4763 0.00 0.0000
scene_03 29.73 0.7996 17.70 0.3379 0.00 0.0000

Public Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 ScoreMicro + gradient 0.849 37.12 0.979 0.93 ✓ Certified Wei et al., ECCV 2025
2 DiffDeconv + gradient 0.846 37.08 0.978 0.92 ✓ Certified Huang et al., NeurIPS 2024
3 DeconvFormer + gradient 0.835 35.79 0.972 0.94 ✓ Certified Chen et al., CVPR 2024
4 Restormer+ + gradient 0.818 35.56 0.971 0.87 ✓ Certified Zamir et al., ICCV 2024
5 Restormer + gradient 0.816 34.39 0.964 0.93 ✓ Certified Zamir et al., CVPR 2022
6 CARE + gradient 0.799 32.85 0.951 0.95 ✓ Certified Weigert et al., Nat. Methods 2018
7 ResUNet + gradient 0.797 34.07 0.961 0.86 ✓ Certified DeCelle et al., Nat. Methods 2021
8 U-Net + gradient 0.785 32.58 0.949 0.9 ✓ Certified Ronneberger et al., MICCAI 2015
9 PnP-DnCNN + gradient 0.749 29.64 0.911 0.94 ✓ Certified Zhang et al., IEEE TIP 2017
10 PnP-FISTA + gradient 0.709 27.9 0.879 0.89 ✓ Certified Bai et al., 2020
11 Wiener Filter + gradient 0.701 27.17 0.862 0.92 ✓ Certified Analytical baseline
12 TV-Deconvolution + gradient 0.693 27.21 0.863 0.88 ✓ Certified Rudin et al., Phys. A 1992
13 Richardson-Lucy + gradient 0.674 25.8 0.826 0.93 ✓ Certified Richardson, JOSA 1972 / Lucy, AJ 1974

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%
Back to Two-Photon