Analysis of gait recognition on constrained scenarios with limited data information
EntityUAM. Departamento de Tecnología Electrónica y de las Comunicaciones
PublisherSpringer Berlin Heidelberg
10.1007/978-3-642-38061-7_23Highlights on Practical Applications of Agents and Multi-Agent Systems: International Workshops of PAAMS. Communications in Computer and Information Science, Volumen 365. Springer, 2013. 231-239
ISBN978-3-642-38060-0 (print); 978-3-642-38061-7 (online)
Funded byThis work has been supported by projects Contexts (S2009/TIC-1485), Bio-Challenge (TEC2009-11186), Bio-Shield (TEC2012-34881) and "Cátedra UAM-Telefonica". Rubén Vera-Rodríguez is supported by a Juan de la Cierva Fellowship from the Spanish MINECO.
ProjectComunidad de Madrid. S2009/TIC-1485/CONTEXTS
SubjectsBiometrics; Gait recognition; Video surveillance; Telecomunicaciones
NoteThe final publication is available at Springer via http://dx.doi.org/10.1007/978-3-642-38061-7_23
Proceedings of International Workshops of Practical Applications of Agents and Multi-Agent Systems (PAAMS), held in 2013, Salamanca (Spain).
Rights© Springer-Verlag Berlin Heidelberg 2013
This paper is focused on the assessment of gait recognition on a constrained scenario, where limited information can be extracted from the gait image sequences. In particular we are interested in assessing the performance of gait images when only the lower part of the body is acquired by the camera and just half of a gait cycle is available (SFootBD database). Thus, various state-of-the-art feature approaches have been followed and applied to the data. Results show that good recognition performance can be achieved using such limited data information for gait biometric. A comparative analysis of the influence of the quantity of data used in the training models has been carried out obtaining results of 8.6% EER for the case of using 10 data samples to train the models, and 5.7% of EER for the case of using 40 data for training. Also, a comparison with a standard and ideal gait database (USF database) is also carried out using similar experimental protocols. In this case 10 data samples are used for training achieving results of 3.6% EER. The comparison with a standard database shows that different feature approaches perform differently for each database, achieving best individual results with MPCA and EGEI methods for the SFootBD and the USF databases respectively
Google Scholar:Vera Rodríguez, Rubén - Gabriel Sanz, Silvia - Fiérrez Aguilar, Julián - Tomé González, Pedro - Ortega García, Javier
This item appears in the following Collection(s)
Showing items related by title, author, creator and subject.