English speech emotion recognition method based on speech recognition

Speech emotion reflects important information other than text content in speech signal, while traditional speech recognition often ignores the emotion of text content, so it is difficult to understand more abundant emotional content from English text. In order to change this situation and get more emotional information from English texts, it is necessary to understand English speech emotion recognition. However, at present, the research on speech emotion recognition technology in China mainly focuses on Chinese, while the research on English speech emotion recognition is relatively few. Therefore, this paper studies English speech emotion recognition. The digital processing of speech signal is based on speech recognition. The digitization of speech signal is the premise of computer processing and analysis of speech signal. The preprocessing of speech signal can also be called front-end processing. The specific steps are: sampling and quantization, pre intensity and windowing. Voice endpoint detection is based on high-order differentiation of volume and waveform. In feature extraction, open smile is selected as the tool to directly extract features, libsvm is selected to establish speech emotion recognition model, and finally an experimental environment is built to verify the design method. The experimental results show that this method can better recognize the emotion of English speech and realize a high degree of human–computer interaction.