Does Pooling Really Matter? An Evaluation on Gait Recognition
dc.contributor.author | dos Santos, Claudio Filipi Goncalves | |
dc.contributor.author | Moreira, Thierry Pinheiro [UNESP] | |
dc.contributor.author | Colombo, Danilo | |
dc.contributor.author | Papa, João Paulo [UNESP] | |
dc.contributor.institution | Universidade Federal de São Carlos (UFSCar) | |
dc.contributor.institution | Universidade Estadual Paulista (Unesp) | |
dc.contributor.institution | Petróleo Brasileiro S.A. – Petrobras | |
dc.date.accessioned | 2020-12-12T02:30:27Z | |
dc.date.available | 2020-12-12T02:30:27Z | |
dc.date.issued | 2019-01-01 | |
dc.description.abstract | Most Convolutional Neural Networks make use of subsampling layers to reduce dimensionality and keep only the most essential information, besides turning the model more robust to rotation and translation variations. One of the most common sampling methods is the one who keeps only the maximum value in a given region, known as max-pooling. In this study, we provide pieces of evidence that, by removing this subsampling layer and changing the stride of the convolution layer, one can obtain comparable results but much faster. Results on the gait recognition task show the robustness of the proposed approach, as well as its statistical similarity to other pooling methods. | en |
dc.description.affiliation | Federal University of São Carlos - UFSCar | |
dc.description.affiliation | State University of Sao Paulo - UNESP | |
dc.description.affiliation | Cenpes Petróleo Brasileiro S.A. – Petrobras | |
dc.description.affiliationUnesp | State University of Sao Paulo - UNESP | |
dc.description.sponsorship | Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) | |
dc.description.sponsorship | Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) | |
dc.description.sponsorshipId | FAPESP: 2013/07375-0 | |
dc.description.sponsorshipId | FAPESP: 2014/12236-1 | |
dc.description.sponsorshipId | FAPESP: 2016/06441-7 | |
dc.description.sponsorshipId | FAPESP: 2017/25908-6 | |
dc.description.sponsorshipId | CNPq: 307066/2017-7 | |
dc.description.sponsorshipId | CNPq: 427968/2018-6 | |
dc.description.sponsorshipId | CNPq: 429003/2018-8 | |
dc.format.extent | 751-760 | |
dc.identifier | http://dx.doi.org/10.1007/978-3-030-33904-3_71 | |
dc.identifier.citation | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), v. 11896 LNCS, p. 751-760. | |
dc.identifier.doi | 10.1007/978-3-030-33904-3_71 | |
dc.identifier.issn | 1611-3349 | |
dc.identifier.issn | 0302-9743 | |
dc.identifier.scopus | 2-s2.0-85075696640 | |
dc.identifier.uri | http://hdl.handle.net/11449/201356 | |
dc.language.iso | eng | |
dc.relation.ispartof | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | |
dc.source | Scopus | |
dc.subject | Convolutional Neural Networks | |
dc.subject | Deep learning | |
dc.subject | Gait recognition | |
dc.title | Does Pooling Really Matter? An Evaluation on Gait Recognition | en |
dc.type | Trabalho apresentado em evento | |
unesp.campus | Universidade Estadual Paulista (Unesp), Faculdade de Ciências, Bauru | pt |
unesp.department | Computação - FC | pt |