Public

Machine Vision / AOI — 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
focus_distance_error -1.0 – 2.0 0.5 mm
lens_distortion_k1 -0.02 – 0.04 0.01 -
exposure_time_drift 9.6 – 10.8 10.2 ms
white_balance_gain 0.98 – 1.04 1.01 -

InverseNet Baseline Scores

Method: CPU_baseline — Mismatch parameter: nominal

Scenario I (Ideal)

13.44 dB

SSIM 0.6187

Scenario II (Mismatch)

11.17 dB

SSIM 0.1544

Scenario III (Oracle)

18.72 dB

SSIM 0.1396

Per-scene breakdown (4 scenes)
Scene PSNR I SSIM I PSNR II SSIM II PSNR III SSIM III
scene_00 13.33 0.6172 12.65 0.1296 18.68 0.1369
scene_01 13.64 0.6232 9.57 0.1826 18.81 0.1378
scene_02 13.41 0.6172 10.74 0.1609 18.65 0.1450
scene_03 13.36 0.6172 11.71 0.1445 18.74 0.1385

Public Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 UniAD + gradient 0.800 33.21 0.954 0.93 ✓ Certified You et al., NeurIPS 2022
2 PatchCore + gradient 0.728 29.11 0.902 0.88 ✓ Certified Roth et al., CVPR 2022
3 PnP-ADMM + gradient 0.696 27.11 0.861 0.9 ✓ Certified Venkatakrishnan et al., 2013
4 Template Match + gradient 0.680 25.88 0.829 0.95 ✓ Certified Brunelli, Template Matching, 2009

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 Machine Vision / AOI