학술논문
Developing Real-Time Streaming Transformer Transducer for Speech Recognition on Large-Scale Dataset
Document Type
Conference
Author
Source
ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Acoustics, Speech and Signal Processing (ICASSP), ICASSP 2021 - 2021 IEEE International Conference on. :5904-5908 Jun, 2021
Subject
Language
ISSN
2379-190X
Abstract
Recently, Transformer based end-to-end models have achieved great success in many areas including speech recognition. However, compared to LSTM models, the heavy computational cost of the Transformer during inference is a key issue to prevent their applications. In this work, we explored the potential of Transformer Transducer (T-T) models for the fist pass decoding with low latency and fast speed on a large-scale dataset. We combine the idea of Transformer- XL and chunk-wise streaming processing to design a streamable Transformer Transducer model. We demonstrate that T-T outperforms the hybrid model, RNN Transducer (RNN-T), and streamable Transformer attention-based encoder-decoder model in the streaming scenario. Furthermore, the runtime cost and latency can be optimized with a relatively small look-ahead.