Browsing through large volumes of spoken audio is known to be a challenging task for end users. One way to facilitate this task is to provide keyphrases extracted from the audio, thus allowing users to quickly get the gist of the audio document or sections of it.
展开▼
