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

The interest in complex deep neural networks for computer vision applicationsis increasing. This leads to the need for improving the interpretablecapabilities of these models. Recent explanation methods present visualizationsof the relevance of pixels from input images, thus enabling the directinterpretation of properties of the input that lead to a specific output. Thesemethods produce maps of pixel importance, which are commonly evaluated byvisual inspection. This means that the effectiveness of an explanation methodis assessed based on human expectation instead of actual feature importance.Thus, in this work we propose an objective measure to evaluate the reliabilityof explanations of deep models. Specifically, our approach is based on changesin the network's outcome resulting from the perturbation of input images in anadversarial way. We present a comparison between widely-known explanationmethods using our proposed approach. Finally, we also propose a straightforwardapplication of our approach to clean relevance maps, creating moreinterpretable maps without any loss in essential explanation (as per ourproposed measure).

Assessing the Reliability of Visual Explanations of Deep Models with Adversarial Perturbations