Voice emotion recognizing method and device, computer device and storage medium

Abstract

The invention discloses a voice emotion recognizing method, a voice emotion recognizing device, a computer device and a computer readable storage medium. The method comprises the following steps: acquiring voice data of a user; determining user attribute information of the user according to the voice data; and according to the user attribute information and the voice data, determining the emotiontype of the user. The method does not depend on the voice recognition technology, the operation of converting voice into a text is not needed, the attribute information of the user is determined through the voice data only, meanwhile, the voice emotion recognizing for the user is realized based on the attribute information, compared with the traditional method for recognizing the emotion of the user through a universal model only, the method provided by the invention realizes the voice emotion recognizing on the user based on the user attributes, therefore, the emotion recognizing effect can be effectively promoted, and the recognizing accuracy is improved.

Description

technical field [0001] The present application relates to the technical field of data processing, and in particular to a speech emotion recognition method, device, computer equipment and computer-readable storage medium. Background technique [0002] Voice interaction is an important interaction mode in the era of artificial intelligence. With the development of the Internet of Things and the promotion of smart hardware, people begin to use voice to communicate, control various smart products through voice, and then recognize the user's emotion according to the user's voice. Users make different emotional feedbacks to make the interactive system more humanized and improve the quality of interaction. [0003] In related technologies, there are mainly three methods for speech emotion recognition. The first method is to use speech recognition technology to convert speech into text, and then use text emotion analysis technology to recognize the emotion in the speech; the second ...

AI Technical Summary

Problems solved by technology

[0004] However, the current technical problem is: the above-mentioned first method, on the one hand, depends on the accuracy of speech recognition. In the dialogue interaction scene, many colloquial expressions are often misrecognized, especially in a relatively noisy environment. On the other hand, many emotional expressions cannot be analyzed simply through text, such as "I want to go shopping." Voice expressions can express emotions such as "anger", "bored" or even "sad", but text cannot. It can be seen that the emotion recognition rate is low; the second method relies on the extraction and selection of various speech features, but the current related technologies cannot extract better emotional speech features, resulting in low accuracy of speech emotion recognition; In the third way, the deep neural network model requires a large amount of voice annotation data, which consumes huge annotation costs, and the voice data in reality is biased and the number of samples is small, which makes it difficult to support the training of complex models.

Images