基于深度学习的语音识别算法的设计与实现

北京理工大学珠海学院2020届本科生毕业设计Design and implementation of speechrecognition algorithm based onDeep LearningABSTRACTSpeech Recognition (Voice Recognit ion)is a technology that allows machines to transformspeech signals into corresponding text or commands by recognizing audio.Speech recognitiontechnology mainly consists of pattern matching recognition,acoustic feature extraction,acoustic model modeling,language model modeling and so on.With the development of deeplearning in the field of machine learning,speech recognition technology has developed by leapsand bounds in recent years.With the development of the Internet,speech recognition has beenwidely used in mobile terminals.On the basis of neural network algorithm and machinetranslation technology in the field of machine learning,this paper designs and implements theapplicat ion of speech recognition algorithm based on deep learning.The algorithm applicationuses the neural network under deep learning as the main architecture,combines CTC,LSTMand other algorithms,uses Pycharm Community Edition as the main development software,andrealizes the algorithm application with Chinese voice answering,Chinese and English text orspeech translat ion functions.Key word:speech recognition;acoustic model;machine learning;deep learning;neuralnetwork algorithm;