Homophone-based Label Smoothing In End-to-end Automatic Speech Recognition

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.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyzheng2020homophone

Related papers

• Right Label Context In End-to-end Training Of Time-synchronous ASR Models (2025)0.00
• Multiple-hypothesis Ctc-based Semi-supervised Adaptation Of End-to-end Speech Recognition (2021)5.84
• Combining Frame-synchronous And Label-synchronous Systems For Speech Recognition (2021)0.00
• Adaptive Frequency Cepstral Coefficients For Word Mispronunciation Detection (2016)5.84
• Audio-attention Discriminative Language Model For ASR Rescoring (2019)9.23
• Alternating Weak Triphone/bpe Alignment Supervision From Hybrid Model Improves End-to-end ASR (2024)0.00
• An Improved Hybrid Ctc-attention Model For Speech Recognition (2018)0.00
• Slimipl: Language-model-free Iterative Pseudo-labeling (2020)10.74