Application of Text Error Correction Algorithm in Railway Speech Recognition

This paper proposes an error detection and error correction algorithm for speech recognition text based on statistics and understanding, in order to optimize speech recognition results. Through experiments, the accuracy rate of the algorithm proposed in this paper is 82.1%, which can effectively correct the errors in the text after speech recognition and realize the application of speech recognition technology in the field of railway signals. This paper has solved the problem of low recognition rate of professional language in the field of railway signal by speech recognition systems.