Exploiting speech for recognizing elderly users to respond to their special needs.

摘要: In this paper we show how to exploit raw speech data gain higher level information about the user in a mobile context. particular introduce an approach for estimation of age and gender using well known machine learning techniques. On basis information, systems like example pedestrian navigation system, can be made adaptive special needs specific group (here elderly). First provide motivation why consider such adaptation as necessary, then outline some strategies that are adequate assistants. The major part is (a) identifying extracting features relevant (b) classifying speaker, treating uncertainty, updating model over time. Finally short outlook on current work.