Application possibilities of hardware implemented hybrid neural networks to support independent life of elderly people

Abstract

With the explosion of new information and communication technologies and new devices, these offer increased opportunities to support everyday life but also have increased the requirements on the expected properties such as adaptability to user needs, behavior and particularities. These requirements are even more needed if the user is elderly, disabled or children. Properties like adaptability or learning capability, self-organization can be ensured by using interfaces that copies biological behavior. Thus, uses of Hybrid Artificial Intelligent Systems can represent key solution for obtaining adaptive interfaces and systems. Modeling such complex systems is some time to computational intensive that's why we have proposed the use of hardware implemented neural networks. Using Field Programmable Gate Arrays (FPGA) for hardware implementation allows parallel implementation of neurons increasing the processing speed. This paper aims to present the method developed by the authors for implementing artificial neural networks, the results obtained and the possibility of use in some applications experimented by the authors support independent life of elderly people. © 2013 Springer-Verlag.