arXiv 2304.110402023-04-20cs.SDcs.CLcs.LGeess.AS

采用机器学习算法的口语情感表达检测

Emotional Expression Detection in Spoken Language Employing Machine Learning Algorithms

University of Information Technology & Sciences（信息技术与科学大学）；Southeast University（东南大学）；Endicott College, Woosong University（又松大学恩迪科特学院）；Daffodil International University（水仙国际大学）；Jahangirnagar University（贾汉吉尔纳加尔大学）

Mehrab Hosain, Most. Yeasmin Arafat, Gazi Zahirul Islam, Jia Uddin, Md. Mobarak Hossain, Fatema Alam

备注：: Journal Pre-print (15 Pages, 9 Figures, 3 Tables)

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

该研究面向口语情感识别，基于 CREMA-D 与 TESS，利用 EMD 和多类声学特征训练 SVM、神经网络、集成学习及 KNN 模型，并验证其性能优势。

摘要：

人类语音具有多种可归类为音高、音色、响度和嗓音色调的特征。在许多事件中可以观察到，人类在说话时会使用不同的声音特质来表达自己的感受。本研究的主要目标是通过使用若干 MATLAB 函数（即频谱描述符、周期性和谐波性）来识别人类的不同情绪，例如愤怒、悲伤、恐惧、中性、厌恶、惊喜和快乐。为完成这项工作，我们分析了人类语音的 CREMA-D（众包情感多模态演员数据）与 TESS（多伦多情感语音集）数据集。音频文件包含具有各种特征（例如嘈杂、快速、缓慢）的数据，从而显著提高 ML（机器学习）模型的效率。EMD（经验模态分解）被用于信号分解过程。随后，通过使用 MFCC、GTCC、频谱质心、滚降点、熵、扩展、通量、谐波比、能量、偏度、平坦度和音频增量等若干技术提取特征。使用一些著名的 ML 模型训练数据，即支持向量机、神经网络、集成学习和 KNN。这些算法在测试数据上的准确率分别为 67.7%、63.3%、61.6% 和 59.0%，在训练数据上的准确率分别为 77.7%、76.1%、99.1% 和 61.2%。我们使用 Matlab 进行了实验，结果表明，与现有的同类工作相比，我们的模型非常突出且更具灵活性。

英文摘要：

There are a variety of features of the human voice that can be classified as pitch, timbre, loudness, and vocal tone. It is observed in numerous incidents that human expresses their feelings using different vocal qualities when they are speaking. The primary objective of this research is to recognize different emotions of human beings such as anger, sadness, fear, neutrality, disgust, pleasant surprise, and happiness by using several MATLAB functions namely, spectral descriptors, periodicity, and harmonicity. To accomplish the work, we analyze the CREMA-D (Crowd-sourced Emotional Multimodal Actors Data) & TESS (Toronto Emotional Speech Set) datasets of human speech. The audio file contains data that have various characteristics (e.g., noisy, speedy, slow) thereby the efficiency of the ML (Machine Learning) models increases significantly. The EMD (Empirical Mode Decomposition) is utilized for the process of signal decomposition. Then, the features are extracted through the use of several techniques such as the MFCC, GTCC, spectral centroid, roll-off point, entropy, spread, flux, harmonic ratio, energy, skewness, flatness, and audio delta. The data is trained using some renowned ML models namely, Support Vector Machine, Neural Network, Ensemble, and KNN. The algorithms show an accuracy of 67.7%, 63.3%, 61.6%, and 59.0% respectively for the test data and 77.7%, 76.1%, 99.1%, and 61.2% for the training data. We have conducted experiments using Matlab and the result shows that our model is very prominent and flexible than existing similar works.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。