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

Lack of interpretability of deep convolutional neural networks (DCNN) is awell-known problem particularly in the medical domain as clinicians wanttrustworthy automated decisions. One way to improve trust is to demonstrate thelocalisation of feature representations with respect to expert labeled regionsof interest. In this work, we investigate the localisation of features learnedvia two varied learning paradigms and demonstrate the superiority of onelearning approach with respect to localisation. Our analysis on medical andnatural datasets show that the traditional end-to-end (E2E) learning strategyhas a limited ability to localise discriminative features across multiplenetwork layers. We show that a layer-wise learning strategy, namely cascadelearning (CL), results in more localised features. Considering localisationaccuracy, we not only show that CL outperforms E2E but that it is a promisingmethod of predicting regions. On the YOLO object detection framework, our bestresult shows that CL outperforms the E2E scheme by $2\%$ in mAP.

Cascade Learning Localises Discriminant Features in Visual Scene Classification