Human-like performance in speech technology could be just around the corner, thanks to a new research project that links three UK universities.
