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

The tremendous potential exhibited by deep learning is often offset byarchitectural and computational complexity, making widespread deployment achallenge for edge scenarios such as mobile and other consumer devices. Totackle this challenge, we explore the following idea: Can we learn generativemachines to automatically generate deep neural networks with efficient networkarchitectures? In this study, we introduce the idea of generative synthesis,which is premised on the intricate interplay between a generator-inquisitorpair that work in tandem to garner insights and learn to generate highlyefficient deep neural networks that best satisfies operational requirements.What is most interesting is that, once a generator has been learned throughgenerative synthesis, it can be used to generate not just one but a largevariety of different, unique highly efficient deep neural networks that satisfyoperational requirements. Experimental results for image classification,semantic segmentation, and object detection tasks illustrate the efficacy ofgenerative synthesis in producing generators that automatically generate highlyefficient deep neural networks (which we nickname FermiNets) with higher modelefficiency and lower computational costs (reaching >10x more efficient andfewer multiply-accumulate operations than several tested state-of-the-artnetworks), as well as higher energy efficiency (reaching >4x improvements inimage inferences per joule consumed on a Nvidia Tegra X2 mobile processor). Assuch, generative synthesis can be a powerful, generalized approach foraccelerating and improving the building of deep neural networks for on-deviceedge scenarios.

FermiNets: Learning generative machines to generate efficient neural networks via generative synthesis