Conv-Transformer Transducer: Low Latency, Low Frame Rate, Streamable End-to-End Speech Recognition
arXiv:2008.05750
Abstract
Transformer has achieved competitive performance against state-of-the-art end-to-end models in automatic speech recognition (ASR), and requires significantly less training time than RNN-based models. The original Transformer, with encoder-decoder architecture, is only suitable for offline ASR. It relies on an attention mechanism to learn alignments, and encodes input audio bidirectionally. The high computation cost of Transformer decoding also limits its use in production streaming systems. To make Transformer suitable for streaming ASR, we explore Transducer framework as a streamable way to learn alignments. For audio encoding, we apply unidirectional Transformer with interleaved convolution layers. The interleaved convolution layers are used for modeling future context which is important to performance. To reduce computation cost, we gradually downsample acoustic input, also with the interleaved convolution layers. Moreover, we limit the length of history context in self-attention to maintain constant computation cost for each decoding step. We show that this architecture, named Conv-Transformer Transducer, achieves competitive performance on LibriSpeech dataset (3.6\% WER on test-clean) without external language models. The performance is comparable to previously published streamable Transformer Transducer and strong hybrid streaming ASR systems, and is achieved with smaller look-ahead window (140~ms), fewer parameters and lower frame rate.
arXiv commentAccepted by INTERSPEECH 2020
References in corpus (7)
- Sequence Transduction with Recurrent Neural Networks
- A Comparative Study on Transformer vs RNN in Speech Applications
- Transformer-based Acoustic Modeling for Hybrid Speech Recognition
- Self-Attention Transducers for End-to-End Speech Recognition
- Stochastic Gradient Methods with Layer-wise Adaptive Moments for Training of Deep Networks
- Generating Wikipedia by Summarizing Long Sequences
- Transformer-Transducer: End-to-End Speech Recognition with Self-Attention