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 learning (DL) defines a data-driven programming paradigm thatautomatically composes the system decision logic from the training data. Incompany with the data explosion and hardware acceleration during the pastdecade, DL achieves tremendous success in many cutting-edge applications.However, even the state-of-the-art DL systems still suffer from quality andreliability issues. It was only until recently that some preliminary progresswas made in testing feed-forward DL systems. In contrast to feed-forward DLsystems, recurrent neural networks (RNN) follow a very different architecturaldesign, implementing temporal behaviors and memory with loops and internalstates. Such stateful nature of RNN contributes to its success in handlingsequential inputs such as audio, natural languages and video processing, butalso poses new challenges for quality assurance.  In this paper, we initiate the very first step towards testing RNN-basedstateful DL systems. We model RNN as an abstract state transition system, basedon which we define a set of test coverage criteria specialized for stateful DLsystems. Moreover, we propose an automated testing framework, DeepCruiser,which systematically generates tests in large scale to uncover defects ofstateful DL systems with coverage guidance. Our in-depth evaluation on astate-of-the-art speech-to-text DL system demonstrates the effectiveness of ourtechnique in improving quality and reliability of stateful DL systems.

DeepCruiser: Automated Guided Testing for Stateful Deep Learning Systems