Publikace UTB
Repozitář publikační činnosti UTB

Dogface detection and localization of dogface’s landmarks

Repozitář DSpace/Manakin

Zobrazit minimální záznam


dc.title Dogface detection and localization of dogface’s landmarks en
dc.contributor.author Vlachynská, Alžběta
dc.contributor.author Komínková Oplatková, Zuzana
dc.contributor.author Tureček, Tomáš
dc.relation.ispartof Advances in Intelligent Systems and Computing
dc.identifier.issn 2194-5357 Scopus Sources, Sherpa/RoMEO, JCR
dc.identifier.isbn 978-3-319-91188-5
dc.date.issued 2019
utb.relation.volume 764
dc.citation.spage 465
dc.citation.epage 476
dc.event.title 7th Computer Science On-line Conference, CSOC 2018
dc.event.sdate 2018-04-25
dc.event.edate 2018-04-28
dc.type conferenceObject
dc.language.iso en
dc.publisher Springer Verlag
dc.identifier.doi 10.1007/978-3-319-91189-2_46
dc.relation.uri https://link.springer.com/chapter/10.1007/978-3-319-91189-2_46
dc.subject Convolutional neural networks en
dc.subject Dogface en
dc.subject Landmark detection en
dc.description.abstract The paper deals with an approach for a reliable dogface detection in an image using the convolutional neural networks. Two detectors were trained on a dataset containing 8351 real-world images of different dog breeds. The first detector achieved the average precision equal to 0.79 while running real-time on single CPU, the second one achieved the average precision equal to 0.98 but more time for processing is necessary. Consequently, the facial landmark detector using the cascade of regressors was proposed based on those, which are commonly used in human face detection. The proposed algorithm is able to detect dog’s eyes, a muzzle, a top of the head and inner bases of the ears with the 0.05 median location error normalized by the inter-ocular distance. The proposed two-step technique – a dogface detection with following facial landmark detector-could be utilized for a dog breeds identification and consequent auto-tagging and image searches. The paper demonstrates a real-world application of the proposed technique – a successful supporting system for taking pictures of dogs facing the camera. © 2019, Springer International Publishing AG, part of Springer Nature. en
utb.faculty Faculty of Applied Informatics
dc.identifier.uri http://hdl.handle.net/10563/1007983
utb.identifier.obdid 43878982
utb.identifier.scopus 2-s2.0-85047851118
utb.identifier.wok 000460247600046
utb.source d-scopus
dc.date.accessioned 2018-07-27T08:47:37Z
dc.date.available 2018-07-27T08:47:37Z
dc.description.sponsorship IC1406, COST, European Cooperation in Science and Technology; CA15140, COST, European Cooperation in Science and Technology; IGA/CebiaTech/2018/003; MSMT-7778/2014, MŠMT, Ministerstvo Školství, Mládeže a Tělovýchovy; LO1303, MŠMT, Ministerstvo Školství, Mládeže a Tělovýchovy; CZ.1.05/2.1.00/03.0089, FEDER, European Regional Development Fund; COST, European Cooperation in Science and Technology
dc.description.sponsorship Ministry of Education, Youth and Sports of the Czech Republic [LO1303 (MSMT-7778/2014)]; European Regional Development Fund under the Project CEBIA-Tech [CZ.1.05/2.1.00/03.0089]; Internal Grant Agency of Tomas Bata University [IGA/CebiaTech/2018/003]; COST (European Cooperation in Science Technology) [CA15140, IC1406]; A. I. Lab at the Faculty of Applied Informatics, Tomas Bata University in Zlin
utb.contributor.internalauthor Vlachynská, Alžběta
utb.contributor.internalauthor Komínková Oplatková, Zuzana
utb.contributor.internalauthor Tureček, Tomáš
utb.fulltext.affiliation Alzbeta Vlachynska, Zuzana Kominkova Oplatkova, Tomas Turecek Faculty of Applied Informatics, Tomas Bata University in Zlin, Nam T.G. Masaryka 5555, 760 01 Zlin, Czech Republic; {vlachynska,oplatkova,tturecek}@utb.cz
utb.fulltext.dates -
utb.scopus.affiliation Faculty of Applied Informatics, Tomas Bata University in Zlin, Nam T.G. Masaryka 5555, Zlin, Czech Republic
utb.fulltext.faculty Faculty of Applied Informatics
utb.fulltext.faculty Faculty of Applied Informatics
utb.fulltext.faculty Faculty of Applied Informatics
Find Full text

Soubory tohoto záznamu

Zobrazit minimální záznam