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

Existing gait recognition methods either directly establish Global FeatureRepresentation (GFR) from original gait sequences or generate Local FeatureRepresentation (LFR) from several local parts. However, GFR tends to neglectlocal details of human postures as the receptive fields become larger in thedeeper network layers. Although LFR allows the network to focus on the detailedposture information of each local region, it neglects the relations amongdifferent local parts and thus only exploits limited local information ofseveral specific regions. To solve these issues, we propose a global-localbased gait recognition network, named GaitGL, to generate more discriminativefeature representations. To be specific, a novel Global and Local ConvolutionalLayer (GLCL) is developed to take full advantage of both global visualinformation and local region details in each layer. GLCL is a dual-branchstructure that consists of a GFR extractor and a mask-based LFR extractor. GFRextractor aims to extract contextual information, e.g., the relationship amongvarious body parts, and the mask-based LFR extractor is presented to exploitthe detailed posture changes of local regions. In addition, we introduce anovel mask-based strategy to improve the local feature extraction capability.Specifically, we design pairs of complementary masks to randomly occludefeature maps, and then train our mask-based LFR extractor on various occludedfeature maps. In this manner, the LFR extractor will learn to fully exploitlocal information. Extensive experiments demonstrate that GaitGL achievesbetter performance than state-of-the-art gait recognition methods. The averagerank-1 accuracy on CASIA-B, OU-MVLP, GREW and Gait3D is 93.6%, 98.7%, 68.0% and63.8%, respectively, significantly outperforming the competing methods. Theproposed method has won the first prize in two competitions: HID 2020 and HID2021.

GaitGL: Learning Discriminative Global-Local Feature Representations for Gait Recognition