Digital phonetics

Professor Vu's research to date has been at the interface between phonetics and speech processing using machine learning methods, particularly Deep Learning. He is interested in using the technological possibilities in the sense of improving broad access for people to information and simplifying processes. Among other things, he has worked on models for speech recognition that can also be used for "smaller" languages with few available resources and do not exclude dialect speakers, for example, from using the technologies. During his time in industrial speech technology research, Professor Vu has demonstrated that his research is of direct relevance to practical applications. He has also found that the speech recognition process should not be separated from the speech understanding process when modelling, as is often the case in current research, but that more basic research is needed. Therefore, he and his research group are conducting research on the connection between the recognition and understanding of spoken language and, furthermore, on methods that enable collaboration between humans and speech processing systems.