Teaching speech emotion recognition method based on dynamic time sequence modeling and multi-scale fusion

Through the teaching speech emotion recognition method of dynamic time series modeling and multi-scale fusion, the shortcomings of emotion recognition under small sample conditions in existing technologies are solved, higher accuracy and robustness are achieved, and it can adapt to the complex changes of emotions in teachers' voices.

CN120727043APending Publication Date: 2025-09-30SHAANXI NORMAL UNIV
0 Cites 3 Cited by

Patent Information

Application Number
CN202511009928.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-22
Publication Date
2025-09-30

AI Technical Summary

Technical Problem

Existing technologies find it difficult to effectively model the short-term dynamic changes and long-term dependent characteristics of emotional expression in teachers' voices under small sample conditions, and are deficient in processing multi-scale temporal structure information, resulting in low accuracy in classroom speech emotion recognition.

Method used

A teaching speech emotion recognition method based on dynamic time series modeling and multi-scale fusion is adopted. Frame-level acoustic features are extracted through the Wav2Vec2.0 pre-training model. Data enhancement and adaptive time shift are performed by combining the AdaShiftFormer learning module and the multi-scale time series fusion module. A speech emotion classification network is constructed, and contrastive learning is used to optimize feature discriminability.

Benefits of technology

It improves the accuracy and robustness of sentiment classification, can more accurately identify emotional changes in teachers' voices under small sample conditions, adapts to the complex and changeable characteristics of classroom speech, and improves the generalization ability of the model.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a teaching speech emotion recognition method based on dynamic time sequence modeling and multi-scale fusion. The method comprises the steps of speech data set preprocessing, speech feature extraction, speech emotion classification network construction, speech emotion classification network training, inputting a test set into the trained speech emotion classification network, and outputting the probability of each type of emotion. According to the method, the direction of the information flow is dynamically adjusted through the adaptive time displacement module, the features of different time scales are extracted by using the multi-scale convolution branch, the modeling capability of the model for a complex time sequence structure and variation data is improved, the feature expression is enhanced, the time sequence features are extracted by using the Wav2Vec2.0 pre-training model, and the time sequence features are extracted by using the Wav2Vec2.0 pre-training model. A speech emotion classification network comprising an AdaShiftFormer learning module and a multi-scale time sequence fusion module is constructed, training is carried out in combination with classification loss and comparison loss, and classification performance is optimized. The method is superior to the prior art in emotion classification accuracy and feature expression ability, and can assist in teacher speech behavior analysis and classroom interaction optimization in an intelligent education scene.
Need to check novelty before this filing date? Find Prior Art