基于声纹情感特征提取的舞蹈机器人自动控制研究  

Automatic control of dance robot based on voice print emotion feature extraction

作　　者：刘芳 LIU Fang(Ankang Vocational Technical College,Ankang Shanxi 725000,China)

机构地区：[1]安康职业技术学院,陕西安康725000

出　　处：《自动化与仪器仪表》2024年第9期276-279,284,共5页Automation & Instrumentation

基　　金：2023年度院级教育教学改革研究项目《将美育融入〈幼儿舞蹈创编〉课程的路径研究》(AZJKY2023004)。

摘　　要：舞蹈机器人的应用场景往往较为嘈杂,为解决这一问题,实现其自动化控制,研究提出一种基于声纹情感特征的机器人模型。首先对机器人自由度和基础架构进行了分析,通过峰值提取实现了声纹特征点的提取,并通过情感分析完成舞蹈动作的自动控制。在实验应用中,结果表明,研究设计模型的类型误判率相比于传统计算机视觉方法下降了25.5%,情绪误判率均值相对下降了10.46%。综上可知,研究提出的基于声纹特征提取与情感分析的舞蹈机器人模型,相对于传统模型具有较为明显的改善作用,能够减少人为调试的需求量。在与其余模型的对比中,也表示研究设计模型更能实现精准的节拍分析,实现音乐和动作相符的舞蹈编排,同时保证其稳定性、真实性和流畅性。The application scene of dancing robot is often noisy.To solve this problem and realize its automatic control,a robot model based on the emotion feature of voice print is proposed.Firstly,the degree of freedom and the infrastructure of the robot are analyzed,the voicing feature points are extracted by peak extraction,and the dance movement is automatically controlled by emotion analysis.In the experimental application,the results show that the type error rate of the research design model is reduced by 25.5%compared with the traditional computer vision method,and the mean emotion error rate is decreased by 10.46%.To sum up,the dance robot model based on voice print feature extraction and sentiment analysis proposed in this study has a relatively obvious improvement role compared with the traditional model,and can reduce the need for human debugging.In comparison with other models,it also shows that the research design model can achieve more accurate beat analysis,achieve the choreography of music and movement,and ensure its stability,authenticity and fluency.

关 键 词：声纹特征 情感分析 舞蹈 机器人 自动化 

分 类 号：TP274.2[自动化与计算机技术—检测技术与自动化装置]