Hidden

Cathodoluminescence (CL) Imaging — Hidden Tier

(5 scenes)

Fully blind server-side evaluation — no data download.

No data downloadable. Algorithm runs server-side on hidden measurements.

Package algorithm as Docker container / Python script. Submit via link.

Containerized algorithm accepting y + H, outputting x_hat + corrected spec.

🔒

True spec hidden — blind evaluation, only ranges available.

Parameter	Spec Range	Unit
beam_current_drift	-0.7 – 2.3	-
collection_efficiency_variation	-2.8 – 9.2	spatial
spectral_calibration_error	-0.28 – 0.92	nm
carbon_contamination	-1.4 – 4.6	signalloss

#	Method	Score	PSNR	SSIM	Consistency	Trust	Source
1	Restormer-CL + gradient	0.762	31.76	0.94	0.84	✓ Certified	Zamir et al., CVPR 2022 (CL adapted)
2	PINN-CL + gradient	0.680	27.63	0.873	0.77	✓ Certified	Raissi et al., J. Comput. Phys. 2019 (CL)
3	SwinIR-CL + gradient	0.667	25.93	0.83	0.88	✓ Certified	Liang et al., ICCV 2021 (CL adapted)
4	CARE-CL + gradient	0.666	26.62	0.849	0.8	✓ Certified	Weigert et al., Nat. Methods 2018 (CL adapted)
5	DiffusionEM + gradient	0.650	25.5	0.818	0.84	✓ Certified	Gao et al., Nat. Methods 2024 (EM adapted)
6	Richardson-Lucy + gradient	0.640	25.3	0.812	0.81	✓ Certified	Richardson, J. Opt. Soc. Am. 1972
7	U-Net-CL + gradient	0.617	24.75	0.794	0.76	✓ Certified	Ronneberger et al., MICCAI 2015 (CL adapted)
8	DnCNN-CL + gradient	0.579	22.56	0.713	0.84	✓ Certified	Zhang et al., IEEE TIP 2017 (CL adapted)
9	Wiener-CL + gradient	0.526	20.5	0.622	0.86	✓ Certified	Castleman, Digital Image Processing, 1996

Scenes: 5

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

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