摘要
本文提出了一种混合特征的语音情感识别算法,该算法利用LSTM深度学习网络对PLP和MFCC特征进行联合学习,以自动提取最优特征表示,从而提升语音情感识别的准确性。在此基础上,本文设计了全新的智能财务客服系统架构。最后,通过实验验证了所提方法在提升智能财务客服系统语音情感识别准确性方面的有效性。
This paper proposes a hybrid feature-based speech emotion recognition algorithm,which utilizes an LSTM deep learning network to jointly learn PLP and MFCC features.This approach aims to automatically extract the optimal feature representations,thereby enhancing the accuracy of speech emotion recognition.Building on this,a novel architecture for an intelligent financial customer service system is designed.Finally,experiments validate the effectiveness of the proposed method in improving the accuracy of speech emotion recognition within the intelligent financial customer service system.
作者
刘晟天
LIU Shengtian(Henan Vocational College of Industry and Informatization,Jiaozuo Henan 454000)
出处
《软件》
2024年第11期7-9,共3页
Software
基金
2023年度中国高校产学研创新基金-正保项目(二期)立项课题“数字化技术推进财务类专业产教融合人才培养体系研究”(2023ZB014)的阶段性研究成果。
关键词
语音情感识别
智能财务
深度学习
speech emotion recognition
smart finance
deep learning
