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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Deep neural networks (DNNs) have been shown lack of robustness for thevulnerability of their classification to small perturbations on the inputs.This has led to safety concerns of applying DNNs to safety-critical domains.Several verification approaches have been developed to automatically prove ordisprove safety properties of DNNs. However, these approaches suffer fromeither the scalability problem, i.e., only small DNNs can be handled, or theprecision problem, i.e., the obtained bounds are loose. This paper improves ona recent proposal of analyzing DNNs through the classic abstract interpretationtechnique, by a novel symbolic propagation technique. More specifically, thevalues of neurons are represented symbolically and propagated forwardly fromthe input layer to the output layer, on top of abstract domains. We show thatour approach can achieve significantly higher precision and thus can prove moreproperties than using only abstract domains. Moreover, we show that the boundsderived from our approach on the hidden neurons, when applied to astate-of-the-art SMT based verification tool, can improve its performance. Weimplement our approach into a software tool and validate it over a few DNNstrained on benchmark datasets such as MNIST, etc.

Analyzing Deep Neural Networks with Symbolic Propagation: Towards Higher Precision and Faster Verification