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

Building invariance to non-meaningful transformations is essential tobuilding efficient and generalizable machine learning models. In practice, themost common way to learn invariance is through data augmentation. There hasbeen recent interest in the development of methods that learn distributions onaugmentation transformations from the training data itself. While suchapproaches are beneficial since they are responsive to the data, they ignorethe fact that in many situations the range of transformations to which a modelneeds to be invariant changes depending on the particular class input belongsto. For example, if a model needs to be able to predict whether an imagecontains a starfish or a dog, we may want to apply random rotations to starfishimages during training (since these do not have a preferred orientation), butwe would not want to do this to images of dogs. In this work we introduce amethod by which we can learn class conditional distributions on augmentationtransformations. We give a number of examples where our methods learn differentnon-meaningful transformations depending on class and further show how ourmethod can be used as a tool to probe the symmetries intrinsic to a potentiallycomplex dataset.

Rotating spiders and reflecting dogs: a class conditional approach to learning data augmentation distributions