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

Depth factorization and quantization have emerged as two of the principalstrategies for designing efficient deep convolutional neural network (CNN)architectures tailored for low-power inference on the edge. However, there isstill little detailed understanding of how different depth factorizationchoices affect the final, trained distributions of each layer in a CNN,particularly in the situation of quantized weights and activations. In thisstudy, we introduce a progressive depth factorization strategy for efficientCNN architecture exploration under quantization constraints. By algorithmicallyincreasing the granularity of depth factorization in a progressive manner, theproposed strategy enables a fine-grained, low-level analysis of layer-wisedistributions. Thus enabling the gain of in-depth, layer-level insights onefficiency-accuracy tradeoffs under fixed-precision quantization. Such aprogressive depth factorization strategy also enables efficient identificationof the optimal depth-factorized macroarchitecture design (which we will referto here as FactorizeNet) based on the desired efficiency-accuracy requirements.

FactorizeNet: Progressive Depth Factorization for Efficient Network Architecture Exploration Under Quantization Constraints