基于传感器技术的情绪采集与电子音乐交互实现  

Emotion acquisition and electronic music interaction realization based on sensor technology

作　　者：贾丹妮[1] JIA Danni(Baoji Vocational and Technical College,Baoji city,Shaanxi Province 721013,China)

机构地区：[1]宝鸡职业技术学院,陕西宝鸡721013

出　　处：《自动化与仪器仪表》2024年第2期242-245,共4页Automation & Instrumentation

基　　金：2022年全省共青团和青年工作研究课题项目(20221802)。

摘　　要：为提高电子音乐的交互性能,提出一种基于压阻式阵列触觉传感器的电子音乐交互方法。该方法通过压阻式阵列触觉传感器主动采集用户触碰信息,并建立触碰压力和不同情绪相对应的数据集;然后通过ELM-ELLA算法对用户情绪进行识别分析;最后在赋予各类歌单情绪标签基础上,通过构建情绪和对应歌单的播放策略,实现电子音乐交互。结果表明,ELM-ELLA算法可利用采集的用户触碰信息分析出用户情绪,且具有持续学习能力,随着算法执行的情绪分析任务的增加,其情绪分析能力也稳步提升;基于传感器技术的电子音乐交互模型输出的音乐,可调节多数人的情绪,基本实现电子音乐交互需求,可提升机器人的个性化服务能力。To improve the interactive performance of electronic music,a method for electronic music interaction based on piezoresistive array tactile sensors is proposed.This method actively collects user touch information through a piezoresistive array tactile sensor,and establishes a dataset corresponding to touch pressure and different emotions;Then,the ELM-ELLA algorithm is used to identify and analyze user emotions;Finally,based on assigning emotional labels to various playlists,electronic music interaction is achieved by constructing a playback strategy for emotions and corresponding playlists.The results show that the ELM-ELLA algorithm can use the collected user touch information to analyze user emotions,and has continuous learning ability.As the number of emotion analysis tasks executed by the algorithm increases,its emotion analysis ability also steadily improves;The electronic music interaction model based on sensor technology outputs music that can regulate the emotions of most people,basically meeting the needs of electronic music interaction,and improving the personalized service ability of the robot.

关 键 词：极限学习机 ELL 人机交互 情绪识别 

分 类 号：TP392[自动化与计算机技术—计算机应用技术]