On the Comparison of Popular End-to-End Models for Large Scale Speech Recognition

Abstract

Recently, there has been a strong push to transition from hybrid models to end-to-end (E2E) models for automatic speech recognition. Currently, there are three promising E2E methods: recurrent neural network transducer (RNN-T), RNN attention-based encoder-decoder (AED), and Transformer-AED. In this study, we conduct an empirical comparison of RNN-T, RNN-AED, and Transformer-AED models, in both non-streaming and streaming modes. We use 65 thousand hours of Microsoft anonymized training data to train these models. As E2E models are more data hungry, it is better to compare their effectiveness with large amount of training data. To the best of our knowledge, no such comprehensive study has been conducted yet. We show that although AED models are stronger than RNN-T in the non-streaming mode, RNN-T is very competitive in streaming mode if its encoder can be properly initialized. Among all three E2E models, transformer-AED achieved the best accuracy in both streaming and non-streaming mode. We show that both streaming RNN-T and transformer-AED models can obtain better accuracy than a highly-optimized hybrid model.

Accepted by Interspeech 2020

References in corpus (5)

• Attention-Based Models for Speech Recognition
• Sequence Transduction with Recurrent Neural Networks
• cuDNN: Efficient Primitives for Deep Learning
• Conformer: Convolution-augmented Transformer for Speech Recognition
• RNN-T For Latency Controlled ASR With Improved Beam Search

Cited by in corpus (6)

• UniSpeech: Unified Speech Representation Learning with Labeled and Unlabeled Data
• Tied & Reduced RNN-T Decoder
• Scaling End-to-End Models for Large-Scale Multilingual ASR
• A Better and Faster End-to-End Model for Streaming ASR
• Improving RNN Transducer Based ASR with Auxiliary Tasks
• Transformer Based Deliberation for Two-Pass Speech Recognition