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

Large-scale face recognition in-the-wild has been recently achieved maturedperformance in many real work applications. However, such systems are built onGPU platforms and mostly deploy heavy deep network architectures. Given ahigh-performance heavy network as a teacher, this work presents a simple andelegant teacher-student learning paradigm, namely ShrinkTeaNet, to train aportable student network that has significantly fewer parameters andcompetitive accuracy against the teacher network. Far apart from priorteacher-student frameworks mainly focusing on accuracy and compression ratiosin closed-set problems, our proposed teacher-student network is proved to bemore robust against open-set problem, i.e. large-scale face recognition. Inaddition, this work introduces a novel Angular Distillation Loss for distillingthe feature direction and the sample distributions of the teacher's hypersphereto its student. Then ShrinkTeaNet framework can efficiently guide the student'slearning process with the teacher's knowledge presented in both intermediateand last stages of the feature embedding. Evaluations on LFW, CFP-FP, AgeDB,IJB-B and IJB-C Janus, and MegaFace with one million distractors havedemonstrated the efficiency of the proposed approach to learn robust studentnetworks which have satisfying accuracy and compact sizes. Our ShrinkTeaNet isable to support the light-weight architecture achieving high performance with99.77% on LFW and 95.64% on large-scale Megaface protocols.

ShrinkTeaNet: Million-scale Lightweight Face Recognition via Shrinking Teacher-Student Networks