|| Online Automatic Speech Recognition With Listen, Attend and Spell Model
||Roger Hsiao, Dogan Can, Tim Ng, Ruchir Travadi, Arnab Ghoshal, Apple Inc, United States|
|Session||SPE-14: Speech Recognition 6: New Algorithms for Sparsity/Efficiency|
|Session Time:||Wednesday, 09 June, 13:00 - 13:45|
|Presentation Time:||Wednesday, 09 June, 13:00 - 13:45|
|| Speech Processing: [SPE-LVCR] Large Vocabulary Continuous Recognition/Search|
|| Click here to watch in the Virtual Conference
|| The Listen, Attend and Spell (LAS) model and other attention-based automatic speech recognition (ASR) models have known limitations when operated in a fully online mode. In this letter, we analyze the online operation of LAS models to demonstrate that these limitations stem from the handling of silence regions and the reliability of online attention mechanism at the edge of input buffers. We propose a novel and simple technique that can achieve fully online recognition while meeting accuracy and latency targets. For the Mandarin dictation task, our proposed approach can achieve a character error rate in online operation that is within 4% relative to an offline LAS model. The proposed online LAS model operates at 12% lower latency relative to a conventional neural network hidden Markov model hybrid of comparable accuracy. We have validated the proposed method through a production scale deployment, which, to the best of our knowledge, is the first such deployment of a fully online LAS model.