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

We propose a method to estimate the uncertainty of the outcome of an imageclassifier on a given input datum. Deep neural networks commonly used for imageclassification are deterministic maps from an input image to an output class.As such, their outcome on a given datum involves no uncertainty, so we mustspecify what variability we are referring to when defining, measuring andinterpreting uncertainty, and attributing "confidence" to the outcome. To thisend, we introduce the Wellington Posterior, which is the distribution ofoutcomes that would have been obtained in response to data that could have beengenerated by the same scene that produced the given image. Since there areinfinitely many scenes that could have generated any given image, theWellington Posterior involves inductive transfer from scenes other than the oneportrayed. We explore the use of data augmentation, dropout, ensembling,single-view reconstruction, and model linearization to compute a WellingtonPosterior. Additional methods include the use of conditional generative modelssuch as generative adversarial networks, neural radiance fields, andconditional prior networks. We test these methods against the empiricalposterior obtained by performing inference on multiple images of the sameunderlying scene. These developments are only a small step towards assessingthe reliability of deep network classifiers in a manner that is compatible withsafety-critical applications and human interpretation.

Scene Uncertainty and the Wellington Posterior of Deterministic Image Classifiers