Human face verification based on multidimensional Polynomial Powers of Sigmoid (PPS)
Nenhuma Miniatura disponível
Data
2008-01-01
Autores
Marar, João Fernando [UNESP]
Coelho, Helder
Título da Revista
ISSN da Revista
Título de Volume
Editor
Insticc-inst Syst Technologies Information Control & Communication
Resumo
In this paper, we described how a multidimensional wavelet neural networks based on Polynomial Powers of Sigmoid (PPS) can be constructed, trained and applied in image processing tasks. In this sense, a novel and uniform framework for face verification is presented. The framework is based on a family of PPS wavelets,generated from linear combination of the sigmoid functions, and can be considered appearance based in that features are extracted from the face image. The feature vectors are then subjected to subspace projection of PPS-wavelet. The design of PPS-wavelet neural networks is also discussed, which is seldom reported in the literature. The Stirling Universitys face database were used to generate the results. Our method has achieved 92 % of correct detection and 5 % of false detection rate on the database.
Descrição
Palavras-chave
artificial neural network, human face verification, Polynomial Powers of Sigmoid (PPS), wavelets functions, PPS-wavelet neural networks, activation functions, feedforward networks
Como citar
Healthinf 2008: Proceedings of The First International Conference on Health Informatics, Vol 2. Setubal: Insticc-inst Syst Technologies Information Control & Communication, p. 99-106, 2008.