A two-parameter generative model of dependent latent feature mixing reproduces natural image correlations in the sparse regime, indicating sparse coding as the appropriate data decomposition.
Sohl-Dickstein, E
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cond-mat.dis-nn 1years
2022 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Generative random latent features models and statistics of natural images
A two-parameter generative model of dependent latent feature mixing reproduces natural image correlations in the sparse regime, indicating sparse coding as the appropriate data decomposition.