English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Decades of psychological research have been aimed at modeling how peoplelearn features and categories. The empirical validation of these theories isoften based on artificial stimuli with simple representations. Recently, deepneural networks have reached or surpassed human accuracy on tasks such asidentifying objects in natural images. These networks learn representations ofreal-world stimuli that can potentially be leveraged to capture psychologicalrepresentations. We find that state-of-the-art object classification networksprovide surprisingly accurate predictions of human similarity judgments fornatural images, but fail to capture some of the structure represented bypeople. We show that a simple transformation that corrects these discrepanciescan be obtained through convex optimization. We use the resultingrepresentations to predict the difficulty of learning novel categories ofnatural images. Our results extend the scope of psychological experiments andcomputational modeling by enabling tractable use of large natural stimulussets.

Evaluating (and improving) the correspondence between deep neural networks and human representations