Human action recognition with 3D convolutional neural networks

Simple item page
dc.contributor.advisorClark, Allanen_ZA
dc.contributor.authorCronje, Fransen_ZA
dc.date.accessioned2015-12-01T09:20:48Z
dc.date.available2015-12-01T09:20:48Z
dc.date.issued2015en_ZA
dc.description.abstractConvolutional neural networks (CNNs) adapt the regular fully-connected neural network (NN) algorithm to facilitate image classification. Recently, CNNs have been demonstrated to provide superior performance across numerous image classification databases including large natural images (Krizhevsky et al., 2012). Furthermore, CNNs are more readily transferable between different image classification problems when compared to common alternatives. The extension of CNNs to video classification is simple and the rationale behind the components of the model are still applicable due to the similarity between image and video data. Previous CNNs have demonstrated good performance upon video datasets, however have not employed methods that have been recently developed and attributed improvements in image classification networks. The purpose of this research to build a CNN model that includes recently developed elements to present a human action recognition model which is up-to-date with current trends in CNNs and current hardware. Focus is applied to ensemble models and methods such as the Dropout technique, developed by Hinton et al. (2012) to reduce overfitting, and learning rate adaptation techniques. The KTH human action dataset is used to assess the CNN model, which, as a widely used benchmark dataset, facilitates the comparison between previous work performed in the literature. Three CNNs are built and trained to provide insight into design choices as well as allow the construction of an ensemble model. The final ensemble model achieved comparative performance to previous CNNs trained upon the KTH data. While the inclusion of new methods to the CNN model did not result in an improvement on previous models, the competitive result provides an alternative combination of architecture and components to other CNN models.en_ZA
dc.identifier.apacitationCronje, F. (2015). <i>Human action recognition with 3D convolutional neural networks</i>. (Thesis). University of Cape Town ,Faculty of Science ,Department of Statistical Sciences. Retrieved from http://hdl.handle.net/11427/15482en_ZA
dc.identifier.chicagocitationCronje, Frans. <i>"Human action recognition with 3D convolutional neural networks."</i> Thesis., University of Cape Town ,Faculty of Science ,Department of Statistical Sciences, 2015. http://hdl.handle.net/11427/15482en_ZA
dc.identifier.citationCronje, F. 2015. Human action recognition with 3D convolutional neural networks. University of Cape Town.en_ZA
dc.identifier.ris TY - Thesis / Dissertation AU - Cronje, Frans AB - Convolutional neural networks (CNNs) adapt the regular fully-connected neural network (NN) algorithm to facilitate image classification. Recently, CNNs have been demonstrated to provide superior performance across numerous image classification databases including large natural images (Krizhevsky et al., 2012). Furthermore, CNNs are more readily transferable between different image classification problems when compared to common alternatives. The extension of CNNs to video classification is simple and the rationale behind the components of the model are still applicable due to the similarity between image and video data. Previous CNNs have demonstrated good performance upon video datasets, however have not employed methods that have been recently developed and attributed improvements in image classification networks. The purpose of this research to build a CNN model that includes recently developed elements to present a human action recognition model which is up-to-date with current trends in CNNs and current hardware. Focus is applied to ensemble models and methods such as the Dropout technique, developed by Hinton et al. (2012) to reduce overfitting, and learning rate adaptation techniques. The KTH human action dataset is used to assess the CNN model, which, as a widely used benchmark dataset, facilitates the comparison between previous work performed in the literature. Three CNNs are built and trained to provide insight into design choices as well as allow the construction of an ensemble model. The final ensemble model achieved comparative performance to previous CNNs trained upon the KTH data. While the inclusion of new methods to the CNN model did not result in an improvement on previous models, the competitive result provides an alternative combination of architecture and components to other CNN models. DA - 2015 DB - OpenUCT DP - University of Cape Town LK - https://open.uct.ac.za PB - University of Cape Town PY - 2015 T1 - Human action recognition with 3D convolutional neural networks TI - Human action recognition with 3D convolutional neural networks UR - http://hdl.handle.net/11427/15482 ER - en_ZA
dc.identifier.urihttp://hdl.handle.net/11427/15482
dc.identifier.vancouvercitationCronje F. Human action recognition with 3D convolutional neural networks. [Thesis]. University of Cape Town ,Faculty of Science ,Department of Statistical Sciences, 2015 [cited yyyy month dd]. Available from: http://hdl.handle.net/11427/15482en_ZA
dc.language.isoengen_ZA
dc.publisher.departmentDepartment of Statistical Sciencesen_ZA
dc.publisher.facultyFaculty of Scienceen_ZA
dc.publisher.institutionUniversity of Cape Town
dc.subject.otherStatistical Sciencesen_ZA
dc.titleHuman action recognition with 3D convolutional neural networksen_ZA
dc.typeMaster Thesis
dc.type.qualificationlevelMasters
dc.type.qualificationnameMScen_ZA
uct.type.filetypeText
uct.type.filetypeImage
uct.type.publicationResearchen_ZA
uct.type.resourceThesisen_ZA
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
thesis_sci_2015_crnfra003-thesis.pdf
Size:
1.89 MB
Format:
Adobe Portable Document Format
Description:
Download
Collections
Masters