Crowd-information-fused speech emotion recognition method and system

Abstract

The invention belongs to the field of artificial intelligence, and particularly relates to a crowd-information-fused speech emotion recognition method and system, and the method comprises the following steps: S1, collecting a user speech signal; S2, preprocessing the voice signal to obtain a Mel spectrum; S3, cutting off mute sections before and after the Mel spectrum; S4, acquiring deep crowd information through a crowd classification network; S5, acquiring Mel spectrum depth information through a Mel spectrum preprocessing network; S6, obtaining fusion information through the SENet fusion features; and S7, obtaining an emotion recognition structure through the classification network. According to the method, crowd information features are fused, so that emotion feature extraction is more accurate; and information fusion is carried out through an SENet channel attention mechanism, so that deep feature extraction can be effectively carried out, and the overall recognition precision is improved.

Description

technical field [0001] The invention belongs to the field of artificial intelligence, and in particular relates to a speech emotion recognition method and system for fusing crowd information. Background technique [0002] Language interaction is one of the earliest ways of human communication, so speech has become the main way for humans to express emotions. With the rise of human-computer interaction, intelligent speech sentiment analysis is becoming more and more important. At present, the main classification of emotions is the seven emotions proposed by Ekman in the last century, namely: neutral, happy, sad, angry, afraid, disgusted, and surprised. [0003] The current mainstream speech emotion recognition methods are based on traditional algorithms or deep learning methods based on simple neural network architectures. The basic process based on the traditional method is: feature extraction of speech, and emotion classification of speech through features. Among them, t...

AI Technical Summary

Problems solved by technology

[0004] In the current technology, because only the shallow information of speech is considered and a simple network structure is used, the recognition rate of emotion recognition is relatively low, and the generalization is relatively poor.

Images