Log in to save to my catalogue

Diversity priors for learning early visual features

Diversity priors for learning early visual features

https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_doaj_primary_oai_doaj_org_article_a368dbaabd784daa8db0f884190d0808

Diversity priors for learning early visual features

About this item

Full title

Diversity priors for learning early visual features

Publisher

Switzerland: Frontiers Research Foundation

Journal title

Frontiers in computational neuroscience, 2015-08, Vol.9, p.104-104

Language

English

Formats

Publication information

Publisher

Switzerland: Frontiers Research Foundation

More information

Scope and Contents

Contents

This paper investigates how utilizing diversity priors can discover early visual features that resemble their biological counterparts. The study is mainly motivated by the sparsity and selectivity of activations of visual neurons in area V1. Most previous work on computational modeling emphasizes selectivity or sparsity independently. However, we a...

Alternative Titles

Full title

Diversity priors for learning early visual features

Identifiers

Primary Identifiers

Record Identifier

TN_cdi_doaj_primary_oai_doaj_org_article_a368dbaabd784daa8db0f884190d0808

Permalink

https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_doaj_primary_oai_doaj_org_article_a368dbaabd784daa8db0f884190d0808

Other Identifiers

ISSN

1662-5188

E-ISSN

1662-5188

DOI

10.3389/fncom.2015.00104

How to access this item