000048415 001__ 48415 000048415 005__ 20200221144151.0 000048415 0247_ $$2doi$$a10.1016/j.imavis.2015.12.006 000048415 0248_ $$2sideral$$a94226 000048415 037__ $$aART-2016-94226 000048415 041__ $$aeng 000048415 100__ $$0(orcid)0000-0001-6324-940X$$aRodriguez, Mario 000048415 245__ $$aA Time-Flexible-Kernel framework for video-based activity recognition 000048415 260__ $$c2016 000048415 5060_ $$aAccess copy available to the general public$$fUnrestricted 000048415 5203_ $$aThis work deals with the challenging task of activity recognition in unconstrained videos. Standard methods are based on video encoding of low-level features using Fisher Vectors or Bag of Features. However, these approaches model every sequence into a single vector with fixed dimensionality that lacks any long-term temporal information, which may be important for recognition, especially of complex activities. This work proposes a novel framework with two main technical novelties: First, a video encoding method that maintains the temporal structure of sequences and second a Time Flexible Kernel that allows comparison of sequences of different lengths and random alignment. Results on challenging benchmarks and comparison to previous work demonstrate the applicability and value of our framework. 000048415 536__ $$9info:eu-repo/grantAgreement/ES/MINECO/TIN2013-45312-R 000048415 540__ $$9info:eu-repo/semantics/openAccess$$aby-nc-nd$$uhttp://creativecommons.org/licenses/by-nc-nd/3.0/es/ 000048415 590__ $$a2.671$$b2016 000048415 591__ $$aCOMPUTER SCIENCE, THEORY & METHODS$$b19 / 104 = 0.183$$c2016$$dQ1$$eT1 000048415 591__ $$aCOMPUTER SCIENCE, SOFTWARE ENGINEERING$$b17 / 106 = 0.16$$c2016$$dQ1$$eT1 000048415 591__ $$aOPTICS$$b26 / 92 = 0.283$$c2016$$dQ2$$eT1 000048415 591__ $$aENGINEERING, ELECTRICAL & ELECTRONIC$$b75 / 260 = 0.288$$c2016$$dQ2$$eT1 000048415 591__ $$aCOMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE$$b37 / 133 = 0.278$$c2016$$dQ2$$eT1 000048415 592__ $$a0.953$$b2016 000048415 593__ $$aComputer Vision and Pattern Recognition$$c2016$$dQ1 000048415 593__ $$aSignal Processing$$c2016$$dQ1 000048415 593__ $$aElectrical and Electronic Engineering$$c2016$$dQ1 000048415 655_4 $$ainfo:eu-repo/semantics/article$$vinfo:eu-repo/semantics/acceptedVersion 000048415 700__ $$0(orcid)0000-0002-0903-5520$$aOrrite, Carlos$$uUniversidad de Zaragoza 000048415 700__ $$0(orcid)0000-0001-7671-7540$$aMedrano, Carlos$$uUniversidad de Zaragoza 000048415 700__ $$aMakris, Dimitrios 000048415 7102_ $$15008$$2785$$aUniversidad de Zaragoza$$bDpto. Ingeniería Electrón.Com.$$cÁrea Tecnología Electrónica 000048415 773__ $$g48-49 (2016), 26-36$$pImage vis. comput.$$tIMAGE AND VISION COMPUTING$$x0262-8856 000048415 8564_ $$s7686556$$uhttps://zaguan.unizar.es/record/48415/files/texto_completo.pdf$$yPostprint 000048415 8564_ $$s48976$$uhttps://zaguan.unizar.es/record/48415/files/texto_completo.jpg?subformat=icon$$xicon$$yPostprint 000048415 909CO $$ooai:zaguan.unizar.es:48415$$particulos$$pdriver 000048415 951__ $$a2020-02-21-13:07:37 000048415 980__ $$aARTICLE