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

Rapid categorization paradigms have a long history in experimentalpsychology: Characterized by short presentation times and speedy behavioralresponses, these tasks highlight the efficiency with which our visual systemprocesses natural object categories. Previous studies have shown thatfeed-forward hierarchical models of the visual cortex provide a good fit tohuman visual decisions. At the same time, recent work in computer vision hasdemonstrated significant gains in object recognition accuracy with increasinglydeep hierarchical architectures. But it is unclear how well these modelsaccount for human visual decisions and what they may reveal about theunderlying brain processes.  We have conducted a large-scale psychophysics study to assess the correlationbetween computational models and human participants on a rapid animal vs.non-animal categorization task. We considered visual representations of varyingcomplexity by analyzing the output of different stages of processing in threestate-of-the-art deep networks. We found that recognition accuracy increaseswith higher stages of visual processing (higher level stages indeedoutperforming human participants on the same task) but that human decisionsagree best with predictions from intermediate stages.  Overall, these results suggest that human participants may rely on visualfeatures of intermediate complexity and that the complexity of visualrepresentations afforded by modern deep network models may exceed those used byhuman participants during rapid categorization.

How Deep is the Feature Analysis underlying Rapid Visual Categorization?