Raw Waveform Encoder With Multi-scale Globally Attentive Locally Recurrent Networks For End-to-end Speech Recognition
2021 · Max W. Y. Lam, Jun Wang, Chao Weng, et al.
Abstract
End-to-end speech recognition generally uses hand-engineered acoustic features as input and excludes the feature extraction module from its joint optimization. To extract learnable and adaptive features and mitigate information loss, we propose a new encoder that adopts globally attentive locally recurrent (GALR) networks and directly takes raw waveform as input. We observe improved ASR performance and robustness by applying GALR on different window lengths to aggregate fine-grain temporal information into multi-scale acoustic features. Experiments are conducted on a benchmark dataset AISHELL-2 and two large-scale Mandarin speech corpus of 5,000 hours and 21,000 hours. With faster speed and comparable model size, our proposed multi-scale GALR waveform encoder achieved consistent character error rate reductions (CERRs) from 7.9% to 28.1% relative over strong baselines, including Conformer and TDNN-Conformer. In particular, our approach demonstrated notable robustness than the traditiona
Authors
(none)
Tags
Stats
Related papers
- Multi-encoder Multi-resolution Framework For End-to-end Speech Recognition (2018)0.00
- End-to-end Speech Recognition From The Raw Waveform (2018)12.81
- Effectiveasr: A Single-step Non-autoregressive Mandarin Speech Recognition Architecture With High Accuracy And Inference Speed (2024)3.58
- Improved Rawnet With Feature Map Scaling For Text-independent Speaker Verification Using Raw Waveforms (2020)14.15
- Rhr-net: A Residual Hourglass Recurrent Neural Network For Speech Enhancement (2019)0.00
- Rawnet: Advanced End-to-end Deep Neural Network Using Raw Waveforms For Text-independent Speaker Verification (2019)15.34
- Gated Recurrent Fusion With Joint Training Framework For Robust End-to-end Speech Recognition (2020)14.55
- Robust Speech Recognition Using Generative Adversarial Networks (2017)11.29