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

Despite the success of a number of recent techniques for visualself-supervised deep learning, there has been limited investigation into therepresentations that are ultimately learned. By leveraging recent advances inthe comparison of neural representations, we explore in this direction bycomparing a contrastive self-supervised algorithm to supervision for simpleimage data in a common architecture. We find that the methods learn similarintermediate representations through dissimilar means, and that therepresentations diverge rapidly in the final few layers. We investigate thisdivergence, finding that these layers strongly fit to their distinct learningobjectives. We also find that the contrastive objective implicitly fits thesupervised objective in intermediate layers, but that the reverse is not true.Our work particularly highlights the importance of the learned intermediaterepresentations, and raises critical questions for auxiliary task design.

Do Self-Supervised and Supervised Methods Learn Similar Visual Representations?