1 paper · 1 filter
Tianyu Huang, Chung Hoon Hong, Carl Wivagg +1
Voice digital assistants must keep up with trending search queries. We rely on a speech recognition model using contextual biasing with a rapidly updated set of entities, instead o…