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/

Zendy Plus

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

Zendy Tools

Zendy Open

Today, Deep Learning (DL) enhances almost every industrial sector, includingsafety-critical areas. The next generation of safety standards will defineappropriate verification techniques for DL-based applications and proposeadequate fault tolerance mechanisms. DL-based applications, like any othersoftware, are susceptible to common random hardware faults such as bit flips,which occur in RAM and CPU registers. Such faults can lead to silent datacorruption. Therefore, it is crucial to develop methods and tools that help toevaluate how DL components operate under the presence of such faults. In thispaper, we introduce two new Fault Injection (FI) frameworks InjectTF andInjectTF2 for TensorFlow 1 and TensorFlow 2, respectively. Both frameworks areavailable on GitHub and allow the configurable injection of random faults intoNeural Networks (NN). In order to demonstrate the feasibility of theframeworks, we also present the results of FI experiments conducted on fourVGG-based Convolutional NNs using two image sets. The results demonstrate howrandom bit flips in the output of particular mathematical operations and layersof NNs affect the classification accuracy. These results help to identify themost critical operations and layers, compare the reliability characteristics offunctionally similar NNs, and introduce selective fault tolerance mechanisms.

Fault Injectors for TensorFlow: Evaluation of the Impact of Random Hardware Faults on Deep CNNs