Conception et réalisation d’un système intelligent d’assistance vocal pour les personnes à mobilité réduite.

Abstract

Millions of people around the world suffer from visual impairment, blindness, old age and other forms of disability. These individuals are subject to many difficulties in moving around. There is a real need to improve their living comfort and minimize their disability, and many solutions are currently on the market to address this issue. Our project aims to provide technical solutions to meet the comfort needs of these people with reduced mobility to move well in their homes. This project is part of the design and development of an intelligent voice assistant based on convolutional neural networks (Deep Learning), to try to make the daily tasks performed by these people automatic and executed by simple voice commands such as 'open the door' or 'turn on the lights', ...etc... To do this, we have developed an intelligent help system based on voice recognition using neural networks, which essentially interacts with electrical equipment to make life simple and user-friendly. Our system is capable of adapting to the needs of each person by : The large choice of voice commands left to the user to control his equipment, so that he is very comfortable when interacting with the system; The interaction of the system with the user who is permanently informed by voice messages announced by the help system. The number of inputs/outputs provided for the connection (equipment and sensors) is important, which offers a multitude of solutions to adapt to the user's environment.