Deep learning networks prefer the human voice – just like ussciencedaily.com1 point·reckel··0 commentsOpen articleSaveView on HN