A Bayesian auxiliary signal derived from noisy data enables learning-based refinement that improves statistical consistency and quality of any denoiser without clean images or precise noise distribution knowledge.
A database of human segmented natural images and its appli- cation to evaluating segmentation algorithms and measuring ecological statistics
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning-based Statistical Refinement for Denoising
A Bayesian auxiliary signal derived from noisy data enables learning-based refinement that improves statistical consistency and quality of any denoiser without clean images or precise noise distribution knowledge.