Fetching the paper…

Direct Acoustics-to-Word Models for English Conversational Speech Recognition · Around