Fast and accurate recurrent neural network acoustic models for speech recognition

Citations Over TimeTop 1% of 2015 papers

Abstract

We have recently shown that deep Long Short-Term Memory (LSTM) recurrent neural networks (RNNs) outperform feed forward deep neural networks (DNNs) as acoustic models for speech recognition. More recently, we have shown that the performance of sequence trained context dependent (CD) hidden Markov model (HMM) acoustic models using such LSTM RNNs can be equaled by sequence trained phone models initialized with connectionist temporal classification (CTC). In this paper, we present techniques that further improve performance of LSTM RNN acoustic models for large vocabulary speech recognition. We show that frame stacking and reduced frame rate lead to more accurate models and faster decoding. CD phone modeling leads to further improvements. We also present initial results for LSTM RNN models outputting words directly.

Related Papers

• → Connectionist temporal classification(2006)5,325 cited
• → Long short-term memory recurrent neural network architectures for large scale acoustic modeling(2014)2,932 cited
• Towards End-To-End Speech Recognition with Recurrent Neural Networks(2014)
• → Context dependent phone models for LSTM RNN acoustic modelling(2015)54 cited