Public

Endoscopy — 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
fiber_coupling -5.0 – 10.0 2.5 %
core_spacing -0.5 – 1.0 0.25 μm
bending_loss -0.3 – 0.6 0.15 dB

InverseNet Baseline Scores

Method: CPU_baseline — Mismatch parameter: nominal

Scenario I (Ideal)

11.40 dB

SSIM 0.3099

Scenario II (Mismatch)

4.64 dB

SSIM 0.0951

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 11.33 0.2391 4.35 0.1045 0.00 0.0000
scene_01 11.16 0.4853 3.99 0.1293 0.00 0.0000
scene_02 11.18 0.2266 5.41 0.0752 0.00 0.0000
scene_03 11.91 0.2887 4.81 0.0714 0.00 0.0000

Public Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 DiffEndo + gradient 0.841 36.77 0.977 0.91 ✓ Certified Gao et al., MICCAI 2024
2 PhysEndo + gradient 0.826 35.58 0.971 0.91 ✓ Certified Chen et al., Med. Image Anal. 2024
3 TransEndo + gradient 0.820 34.89 0.967 0.92 ✓ Certified Wang et al., Med. Image Anal. 2022
4 SwinEndo + gradient 0.813 34.66 0.966 0.9 ✓ Certified Li et al., IEEE TMI 2023
5 EndoSLAM-Net + gradient 0.791 32.53 0.948 0.93 ✓ Certified Ozyoruk et al., Med. Image Anal. 2021
6 DnCNN-Endo + gradient 0.729 29.15 0.903 0.88 ✓ Certified Zhang et al., IEEE TIP 2017
7 BM3D-Endo + gradient 0.683 26.69 0.85 0.88 ✓ Certified Dabov et al., IEEE TIP 2007
8 CLAHE-Endo + gradient 0.663 25.34 0.813 0.92 ✓ Certified Zuiderveld, Graphics Gems IV 1994
9 Histogram-Eq + gradient 0.597 22.41 0.707 0.95 ✓ Certified Gonzalez & Woods, Digital Image Processing 2002

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%
Back to Endoscopy