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

Deep neural networks paved the way for significant improvements in imagevisual categorization during the last years. However, even though the tasks arehighly varying, differing in complexity and difficulty, existing solutionsmostly build on the same architectural decisions. This also applies to theselection of activation functions (AFs), where most approaches build onRectified Linear Units (ReLUs). In this paper, however, we show that the choiceof a proper AF has a significant impact on the classification accuracy, inparticular, if fine, subtle details are of relevance. Therefore, we propose tomodel the degree of absence and the presence of features via the AF by usingpiece-wise linear functions, which we refer to as L*ReLU. In this way, we canensure the required properties, while still inheriting the benefits in terms ofcomputational efficiency from ReLUs. We demonstrate our approach for the taskof Fine-grained Visual Categorization (FGVC), running experiments on sevendifferent benchmark datasets. The results do not only demonstrate superiorresults but also that for different tasks, having different characteristics,different AFs are selected.

L*ReLU: Piece-wise Linear Activation Functions for Deep Fine-grained Visual Categorization