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

Machine learning at the edge offers great benefits such as increased privacyand security, low latency, and more autonomy. However, a major challenge isthat many devices, in particular edge devices, have very limited memory, weakprocessors, and scarce energy supply. We propose a hybrid hardware-softwareframework that has the potential to significantly reduce the computationalcomplexity and memory requirements of on-device machine learning. In the firststep, inspired by compressive sensing, data is collected in compressed formsimultaneously with the sensing process. Thus this compression happens alreadyat the hardware level during data acquisition. But unlike in compressivesensing, this compression is achieved via a projection operator that isspecifically tailored to the desired machine learning task. The second stepconsists of a specially designed and trained deep network. As concrete examplewe consider the task of image classification, although the proposed frameworkis more widely applicable. An additional benefit of our approach is that it canbe easily combined with existing on-device techniques. Numerical simulationsillustrate the viability of our method.

What Happens on the Edge, Stays on the Edge: Toward Compressive Deep Learning