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

Dynamic neural networks (DyNNs) have become viable techniques to enableintelligence on resource-constrained edge devices while maintainingcomputational efficiency. In many cases, the implementation of DyNNs can besub-optimal due to its underlying backbone architecture being developed at thedesign stage independent of both: (i) the dynamic computing features, e.g.early exiting, and (ii) the resource efficiency features of the underlyinghardware, e.g., dynamic voltage and frequency scaling (DVFS). Addressing this,we present HADAS, a novel Hardware-Aware Dynamic Neural Architecture Searchframework that realizes DyNN architectures whose backbone, early exitingfeatures, and DVFS settings have been jointly optimized to maximize performanceand resource efficiency. Our experiments using the CIFAR-100 dataset and adiverse set of edge computing platforms have seen HADAS dynamic models achieveup to 57% energy efficiency gains compared to the conventional dynamic oneswhile maintaining the desired level of accuracy scores. Our code is availableat https://github.com/HalimaBouzidi/HADAS

HADAS: Hardware-Aware Dynamic Neural Architecture Search for Edge Performance Scaling