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

A number of studies have demonstrated the efficacy of deep learningconvolutional neural network (CNN) models for ocular-based user recognition inmobile devices. However, these high-performing networks have enormous space andcomputational complexity due to the millions of parameters and computationsinvolved. These requirements make the deployment of deep learning models toresource-constrained mobile devices challenging. To this end, only a handful ofstudies based on knowledge distillation and patch-based models have beenproposed to obtain compact size CNN models for ocular recognition in the mobileenvironment. In order to further advance the state-of-the-art, this study forthe first time evaluates five neural network pruning methods and compares themwith the knowledge distillation method for on-device CNN inference and mobileuser verification using ocular images. Subject-independent analysis on VISOBand UPFR-Periocular datasets suggest the efficacy of layerwise magnitude-basedpruning at a compression rate of 8 for mobile ocular-based authentication usingResNet50 as the base model. Further, comparison with the knowledge distillationsuggests the efficacy of knowledge distillation over pruning methods in termsof verification accuracy and the real-time inference measured as deep featureextraction time on five mobile devices, namely, iPhone 6, iPhone X, iPhone XR,iPad Air 2 and iPad 7th Generation.

Compact CNN Models for On-device Ocular-based User Recognition in Mobile Devices