A Study on the improvement of chinese automatic speech recognition accuracy using a lexicon

In this paper, in order to improve the error rate that occurs when Automatic Speech Recognition (ASR), one of the technologies widely applied in the field of speech recognition, is applied to Chinese, the study results on how to improve the Chinese recognition error rate by proposing a model to be added are described. As a result of testing by applying the xlsr-53 data set based on the Wav2vec2.0 model, it was confirmed that the Chinese recognition rate was improved.