Desktop Voice Assistant for Elderly People using Feed-forward Neural Network for Intent Recognition

Technology has deeply embedded itself in our daily lives, making it easier to carry out tasks. However, the use of technology can be complicated for some like elderly people, cognitively and visually impaired individuals, illiterate, or people with limited education. A proposed desktop voice assistant for elderly people using a feed-forward neural network for intent recognition to simplify complex tasks and make technology more accessible for those who find it difficult. The assistant is a companion to assist, automate, and make it easier for older people to use computers. It aids them in understanding, responding to user needs, and assisting with task management and data retrieval. The Feed-forward Neural Network trains the pre-processed data for voice-to-intent recognition. The proposed and developed desktop assistant does voice recognition with an accuracy of around 78.5 percent, external APIs are used to enhance user experience simplify interactions, and boost productivity. The assistant's configurable personality, which is tailored to the user's preferences, makes it stand out. It provides a high level of interactivity, enabling it to be used quickly and is very flexible regarding individual requirements.