Automatic Spelling Correction with Transformer for CTC-based End-to-End Speech Recognition

Abstract

Connectionist Temporal Classification (CTC) based end-to-end speech recognition system usually need to incorporate an external language model by using WFST-based decoding in order to achieve promising results. This is more essential to Mandarin speech recognition since it owns a special phenomenon, namely homophone, which causes a lot of substitution errors. The linguistic information introduced by language model will help to distinguish these substitution errors. In this work, we propose a transformer based spelling correction model to automatically correct errors especially the substitution errors made by CTC-based Mandarin speech recognition system. Specifically, we investigate using the recognition results generated by CTC-based systems as input and the ground-truth transcriptions as output to train a transformer with encoder-decoder architecture, which is much similar to machine translation. Results in a 20,000 hours Mandarin speech recognition task show that the proposed spelling correction model can achieve a CER of 3.41%, which results in 22.9% and 53.2% relative improvement compared to the baseline CTC-based systems decoded with and without language model respectively.

6pages, 5 figures

References in corpus (7)

• SGDR: Stochastic Gradient Descent with Warm Restarts
• End-to-end Continuous Speech Recognition using Attention-based Recurrent NN: First Results
• EESEN: End-to-End Speech Recognition using Deep RNN Models and WFST-based Decoding
• Feedforward Sequential Memory Networks: A New Structure to Learn Long-term Dependency
• ASR Context-Sensitive Error Correction Based on Microsoft N-Gram Dataset
• Deep-FSMN for Large Vocabulary Continuous Speech Recognition
• A spelling correction model for end-to-end speech recognition

Cited by in corpus (4)

• N-best T5: Robust ASR Error Correction using Multiple Input Hypotheses and Constrained Decoding Space
• Reduce and Reconstruct: ASR for Low-Resource Phonetic Languages
• Hybrid phonetic-neural model for correction in speech recognition systems
• Cascade RNN-Transducer: Syllable Based Streaming On-device Mandarin Speech Recognition with a Syllable-to-Character Converter