Voice emotion recognition method and device, equipment and storage medium

Abstract

The invention discloses a voice emotion recognition method and device, equipment and a storage medium. The method comprises the steps of obtaining a to-be-recognized voice signal; extracting at least one emotion feature parameter in the voice signal; inputting the at least one emotion feature parameter into a predetermined first model to obtain an emotion state sequence; and inputting the emotion state sequence into a predetermined second model to obtain an emotion state recognition result which corresponds to the voice signal and meets a preset recognition condition. Thus, the first model is used for dynamic time sequence modeling, the second model is used for classification decision making, the first model and the second model are combined for voice emotion recognition, and the emotion recognition rate of voice signals is increased.

Description

technical field [0001] The invention relates to the field of voice recognition, in particular to a voice emotion recognition method, device, equipment and storage medium. Background technique [0002] With the rapid development of information technology and the increasing dependence of human beings on computers, the ability of human-computer interaction has been paid more and more attention by researchers. In fact, the problems that need to be solved in human-computer interaction are consistent with the important factors in human-to-human communication, and the most important thing is the ability of "voice emotion recognition". [0003] At present, before performing speech emotion recognition operations, researchers will input speech data to be recognized into a pre-established emotion recognition model for speech emotion recognition. However, the emotional data for establishing an emotional recognition model often comes from a noise-free laboratory, and expressive actors r...

AI Technical Summary

Problems solved by technology

However, the emotional data for building an emotional recognition model often comes from a noise-free laboratory, and expressive actors read different emotional color sentences. Such emotional data has a large deviation from the real emotional data.

Therefore, when using the above established emotion recognition model for speech emotion recognition again, there is a large difference between the predicted speech emotion recognition result and the real speech emotion recognition result, that is, the speech emotion recognition rate is low

Images