Adding Connectionist Temporal Summarization Into Conformer To Improve Its Decoder Efficiency For Speech Recognition
2022 · Nick J. C. Wang, Zongfeng Quan, Shaojun Wang, et al.
Abstract
The Conformer model is an excellent architecture for speech recognition modeling that effectively utilizes the hybrid losses of connectionist temporal classification (CTC) and attention to train model parameters. To improve the decoding efficiency of Conformer, we propose a novel connectionist temporal summarization (CTS) method that reduces the number of frames required for the attention decoder fed from the acoustic sequences generated by the encoder, thus reducing operations. However, to achieve such decoding improvements, we must fine-tune model parameters, as cross-attention observations are changed and thus require corresponding refinements. Our final experiments show that, with a beamwidth of 4, the LibriSpeech's decoding budget can be reduced by up to 20% and for FluentSpeech data it can be reduced by 11%, without losing ASR accuracy. An improvement in accuracy is even found for the LibriSpeech "test-other" set. The word error rate (WER) is reduced by 6% relative at the beam wi
Authors
(none)
Tags
Stats
Related papers
- Efficient Conformer: Progressive Downsampling And Grouped Attention For Automatic Speech Recognition (2021)13.79
- Blank Collapse: Compressing CTC Emission For The Faster Decoding (2022)0.00
- An Improved Hybrid Ctc-attention Model For Speech Recognition (2018)0.00
- Fast Conformer With Linearly Scalable Attention For Efficient Speech Recognition (2023)14.47
- Df-conformer: Integrated Architecture Of Conv-tasnet And Conformer Using Linear Complexity Self-attention For Speech Enhancement (2021)11.29
- Stateful Conformer With Cache-based Inference For Streaming Automatic Speech Recognition (2023)8.60
- Order-preserving Abstractive Summarization For Spoken Content Based On Connectionist Temporal Classification (2017)3.58
- Comparison Of Decoding Strategies For CTC Acoustic Models (2017)10.48