Template matching on pure noise yields asymptotic convergence of maximum-likelihood class means and 3D reconstructions to deterministic noise-dependent transforms of the user templates, producing structure-from-noise artifacts.
Probability: theory and examples, volume 49
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2025 2representative citing papers
Derives closed-form optimal loss for unified diffusion models, provides variance-controlled estimators, and shows improved diagnosis, training schedules, and power-law scaling after subtracting the optimal value.
citing papers explorer
-
Structure from Noise: Confirmation Bias in Particle Picking in Structural Biology
Template matching on pure noise yields asymptotic convergence of maximum-likelihood class means and 3D reconstructions to deterministic noise-dependent transforms of the user templates, producing structure-from-noise artifacts.
-
Diagnosing and Improving Diffusion Models by Estimating the Optimal Loss Value
Derives closed-form optimal loss for unified diffusion models, provides variance-controlled estimators, and shows improved diagnosis, training schedules, and power-law scaling after subtracting the optimal value.