Homophone-based Label Smoothing in End-to-End Automatic Speech Recognition

2020/04/07 by Yi Zheng, Zheng, Yi, Xianjie Yang +3
Computer Science · #Audio and Speech Processing (eess.AS) #Computation and Language (cs.CL) #FOS: Computer and information sciences #FOS: Electrical engineering #Music and Audio Processing #Sound (cs.SD) #Speech Recognition and Synthesis #Speech and Audio Processing #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2004.03437

openalex publication_date 2020/04/07 · openalex created_date 2020/04/17 · openalex updated_date 2026/07/28

Abstract

A new label smoothing method that makes use of prior knowledge of a language at human level, homophone, is proposed in this paper for automatic speech recognition (ASR). Compared with its forerunners, the proposed method uses pronunciation knowledge of homophones in a more complex way. End-to-end ASR models that learn acoustic model and language model jointly and modelling units of characters are necessary conditions for this method. Experiments with hybrid CTC sequence-to-sequence model show that the new method can reduce character error rate (CER) by 0.4% absolutely.

Citations

• Attention-Based Models for Speech Recognition
• Long Short-Term Memory Based Recurrent Neural Network Architectures for Large Vocabulary Speech Recognition
• Distilling the Knowledge in a Neural Network
• ESPnet: End-to-End Speech Processing Toolkit

Related

• PAC: Pronunciation-Aware Contextualized Large Language Model-based Automatic Speech Recognition
• A systematic comparison of grapheme-based vs. phoneme-based label units for encoder-decoder-attention models
• Semi-supervised acoustic modelling for five-lingual code-switched ASR using automatically-segmented soap opera speech
• Leveraging Language ID to Calculate Intermediate CTC Loss for Enhanced Code-Switching Speech Recognition
• End-to-End Speech Recognition: A Survey