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 ability to customize a trained Deep Neural Network (DNN) locally usinguser-specific data may greatly enhance user experiences, reduce developmentcosts, and protect user's privacy. In this work, we propose to incorporate anovel Mixture of Experts (MOE) approach to accomplish this goal. Thisarchitecture comprises of a Global Expert (GE), a Local Expert (LE) and aGating Network (GN). The GE is a trained DNN developed on a large trainingdataset representative of many potential users. After deployment on an embeddededge device, GE will be subject to customized, user-specific data (e.g., accentin speech) and its performance may suffer. This problem may be alleviated bytraining a local DNN (the local expert, LE) on a small size customized trainingdata to correct the errors made by GE. A gating network then will be trained todetermine whether an incoming data should be handled by GE or LE. Since thecustomized dataset is in general very small, the cost of training LE and GNwould be much lower than that of re-training of GE. The training of LE and GNthus can be performed at local device, properly protecting the privacy ofcustomized training data. In this work, we developed a prototype MOEarchitecture for handwritten alphanumeric character recognition task. We useEMNIST as the generic dataset, LeNet5 as GE, and handwritings of 10 users asthe customized dataset. We show that with the LE and GN, the classificationaccuracy is significantly enhanced over the customized dataset with almost nodegradation of accuracy over the generic dataset. In terms of energy andnetwork size, the overhead of LE and GN is around 2.5% compared to those of GE.

A Mixture of Expert Approach for Low-Cost Customization of Deep Neural Networks