Recognizing human actions from noisy videos via multiple instance learning

dc.contributor.authorşener, Fadimeen_US
dc.contributor.authorSamet, Nerminen_US
dc.contributor.authorDuygulu, Pınaren_US
dc.contributor.authorIkizler-Cinbis, N.en_US
dc.coverage.spatialHaspolat, Turkeyen_US
dc.date.accessioned2016-02-08T12:07:54Z
dc.date.available2016-02-08T12:07:54Z
dc.date.issued2013en_US
dc.departmentDepartment of Computer Engineeringen_US
dc.descriptionDate of Conference: 24-26 April 2013en_US
dc.description.abstractIn this work, we study the task of recognizing human actions from noisy videos and effects of noise to recognition performance and propose a possible solution. Datasets available in computer vision literature are relatively small and could include noise due to labeling source. For new and relatively big datasets, noise amount would possible increase and the performance of traditional instance based learning methods is likely to decrease. In this work, we propose a multiple instance learning-based solution in case of an increase in noise. For this purpose, each video is represented with spatio-temporal features, then bag-of-words method is applied. Then, using support vector machines (SVM), both instance-based learning and multiple instance learning classifiers are constructed and compared. The classification results show that multiple instance learning classifiers has better performance than instance based learning counterparts on noisy videos. © 2013 IEEE.en_US
dc.description.provenanceMade available in DSpace on 2016-02-08T12:07:54Z (GMT). No. of bitstreams: 1 bilkent-research-paper.pdf: 70227 bytes, checksum: 26e812c6f5156f83f0e77b261a471b5a (MD5) Previous issue date: 2013en
dc.identifier.doi10.1109/SIU.2013.6531431en_US
dc.identifier.urihttp://hdl.handle.net/11693/27998en_US
dc.language.isoTurkishen_US
dc.publisherIEEEen_US
dc.relation.isversionofhttp://dx.doi.org/10.1109/SIU.2013.6531431en_US
dc.source.title2013 21st Signal Processing and Communications Applications Conference (SIU)en_US
dc.subjectData noiseen_US
dc.subjectHuman action recognitionen_US
dc.subjectMultiple instance learningen_US
dc.subjectVideo understandingen_US
dc.subjectClassification resultsen_US
dc.subjectData noiseen_US
dc.subjectHuman-action recognitionen_US
dc.subjectInstance based learningen_US
dc.subjectMultiple instance learningen_US
dc.subjectRecognition performanceen_US
dc.subjectSpatio-temporal featuresen_US
dc.subjectVideo understandingen_US
dc.subjectGesture recognitionen_US
dc.subjectSignal processingen_US
dc.subjectSupport vector machinesen_US
dc.titleRecognizing human actions from noisy videos via multiple instance learningen_US
dc.title.alternativeGürültü içeren videolardan insan hareketlerinin çoklu örnekle ö̌grenme ile taninmasien_US
dc.typeConference Paperen_US

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
Recognizing human actions from noisy videos via multiple instance learning [Gürültü içeren videolardan insan hareketlerinin çoklu örnekle ö̌grenme ile taninmasi].pdf
Size:
4.92 MB
Format:
Adobe Portable Document Format
Description:
Full printable version