Vis enkel innførsel

dc.contributor.authorLeong, Mei Chee
dc.contributor.authorPrasad, Dilip K.
dc.contributor.authorLee, Yong Tsui
dc.contributor.authorLin, Feng
dc.date.accessioned2020-06-26T10:43:37Z
dc.date.available2020-06-26T10:43:37Z
dc.date.issued2020-01-12
dc.description.abstractThis paper introduces a fusion convolutional architecture for efficient learning of spatio-temporal features in video action recognition. Unlike 2D convolutional neural networks (CNNs), 3D CNNs can be applied directly on consecutive frames to extract spatio-temporal features. The aim of this work is to fuse the convolution layers from 2D and 3D CNNs to allow temporal encoding with fewer parameters than 3D CNNs. We adopt transfer learning from pre-trained 2D CNNs for spatial extraction, followed by temporal encoding, before connecting to 3D convolution layers at the top of the architecture. We construct our fusion architecture, semi-CNN, based on three popular models: VGG-16, ResNets and DenseNets, and compare the performance with their corresponding 3D models. Our empirical results evaluated on the action recognition dataset UCF-101 demonstrate that our fusion of 1D, 2D and 3D convolutions outperforms its 3D model of the same depth, with fewer parameters and reduces overfitting. Our semi-CNN architecture achieved an average of 16–30% boost in the top-1 accuracy when evaluated on an input video of 16 frames.en_US
dc.identifier.citationLeong, Prasad DK, Lee, Lin F. Semi-CNN architecture for effective spatio-temporal Learning in action recognition. Applied Sciences. 2020;10(557)en_US
dc.identifier.cristinIDFRIDAID 1815560
dc.identifier.doi10.3390/app10020557
dc.identifier.issn2076-3417
dc.identifier.urihttps://hdl.handle.net/10037/18670
dc.language.isoengen_US
dc.publisherMDPIen_US
dc.relation.journalApplied Sciences
dc.rights.accessRightsopenAccessen_US
dc.rights.holderCopyright 2020 The Author(s)en_US
dc.subjectVDP::Technology: 500::Information and communication technology: 550en_US
dc.subjectVDP::Teknologi: 500::Informasjons- og kommunikasjonsteknologi: 550en_US
dc.titleSemi-CNN architecture for effective spatio-temporal Learning in action recognitionen_US
dc.type.versionpublishedVersionen_US
dc.typeJournal articleen_US
dc.typeTidsskriftartikkelen_US
dc.typePeer revieweden_US


Tilhørende fil(er)

Thumbnail

Denne innførselen finnes i følgende samling(er)

Vis enkel innførsel