An Efficient SMOTE-Based Deep Learning Model for Voice Pathology Detection

被引:8
|
作者
Lee, Ji-Na [1 ]
Lee, Ji-Yeoun [2 ]
机构
[1] Seokyeong Univ, Div Global Business Languages, Seoul 02173, South Korea
[2] Eulji Univ, Dept Bigdata Med Convergence, 553 Sanseong daero, Seongnam Si 13135, South Korea
来源
APPLIED SCIENCES-BASEL | 2023年 / 13卷 / 06期
基金
新加坡国家研究基金会;
关键词
pathological voice; disordered voice; imbalanced learning; voice pathology classification; SMOTE; ADASYN; Borderline-SMOTE; deep learning; intelligent medical diagnosis system; DISEASE DETECTION; IMBALANCED DATA;
D O I
10.3390/app13063571
中图分类号
O6 [化学];
学科分类号
0703 ;
摘要
The Saarbruecken Voice Database (SVD) is a public database used by voice pathology detection systems. However, the distributions of the pathological and normal voice samples show a clear class imbalance. This study aims to develop a system for the classification of pathological and normal voices that uses efficient deep learning models based on various oversampling methods, such as the adaptive synthetic sampling (ADASYN), synthetic minority oversampling technique (SMOTE), and Borderline-SMOTE directly applied to feature parameters. The suggested combinations of oversampled linear predictive coefficients (LPCs), mel-frequency cepstral coefficients (MFCCs), and deep learning methods can efficiently classify pathological and normal voices. The balanced datasets from ADASYN, SMOTE, and Borderline-SMOTE are used to validate and evaluate the various deep learning models. The experiments are conducted using model evaluation metrics such as the recall, specificity, G, and F1 value. The experimental results suggest that the proposed voice pathology detection (VPD) system integrating the LPCs oversampled by the SMOTE and a convolutional neural network (CNN) can effectively yield the highest accuracy at 98.89% when classifying pathological and normal voices. Finally, the performances of oversampling algorithms such as the ADASYN, SMOTE, and Borderline-SMOTE are discussed. Furthermore, the performance of SMOTE is superior to conventional imbalanced data oversampling algorithms, and it can be used to diagnose pathological signals in real-world applications.
引用
收藏
页数:16
相关论文
共 50 条
  • [31] Deep Learning Based Pathology Detection for Smart Connected Healthcares
    Hossain, M. Shamim
    Muhammad, Ghulam
    IEEE NETWORK, 2020, 34 (06): : 120 - 125
  • [32] AN EFFICIENT TRANSFORMER-BASED MODEL FOR VOICE ACTIVITY DETECTION
    Zhao, Yifei
    Champagne, Benoit
    2022 IEEE 32ND INTERNATIONAL WORKSHOP ON MACHINE LEARNING FOR SIGNAL PROCESSING (MLSP), 2022,
  • [33] Voice pathology detection based on the modified voice contour and SVM
    Ali, Zulfiqar
    Alsulaiman, Mansour
    Elamvazuthi, Irraivan
    Muhammad, Ghulam
    Mesallam, Tamer A.
    Farahat, Mohamed
    Malki, Khalid H.
    BIOLOGICALLY INSPIRED COGNITIVE ARCHITECTURES, 2016, 15 : 10 - 18
  • [34] Voice Pathology Detection Using Machine Learning Technique
    AL-Dhief, Fahad Taha
    Mu, Nurul
    Abd Malik, Nik Noordini Nik
    Sabri, Naseer
    Baki, Marina Mat
    Albadr, Musatafa Abbas Abbood
    Abbas, Aymen Fadhil
    Hussein, Yaqdhan Mahmood
    Mohammed, Mazin Abed
    2020 IEEE 5TH INTERNATIONAL SYMPOSIUM ON TELECOMMUNICATION TECHNOLOGIES (ISTT), 2020, : 99 - 104
  • [35] AUC optimization for deep learning-based voice activity detection
    Xiao-Lei Zhang
    Menglong Xu
    EURASIP Journal on Audio, Speech, and Music Processing, 2022
  • [36] AUC optimization for deep learning-based voice activity detection
    Zhang, Xiao-Lei
    Xu, Menglong
    EURASIP JOURNAL ON AUDIO SPEECH AND MUSIC PROCESSING, 2022, 2022 (01)
  • [37] MARNet: An Efficient Two-Stage Intrusion Detection Model Based on Deep Learning
    Wu, Jiang
    Fu, Qiang
    Wang, Liang
    IEEE ACCESS, 2025, 13 : 2377 - 2388
  • [38] An Efficient Real Time Model For Credit Card Fraud Detection Based On Deep Learning
    Abakarim, Youness
    Lahby, Mohamed
    Attioui, Abdelbaki
    PROCEEDINGS OF THE 12TH INTERNATIONAL CONFERENCE ON INTELLIGENT SYSTEMS: THEORIES AND APPLICATIONS (SITA'18), 2018,
  • [39] Deep Learning Approaches for Voice Activity Detection
    Wang, Mantao
    Huang, Qiang
    Zhang, Jie
    Li, Zhiyong
    Pu, Haibo
    Lei, Jinglan
    Wang, Lanjing
    CYBER SECURITY INTELLIGENCE AND ANALYTICS, 2020, 928 : 816 - 826
  • [40] Pedestrian Detection Based on Deep Learning Model
    Li, Hailong
    Wu, Zhendong
    Zhang, Jianwu
    2016 9TH INTERNATIONAL CONGRESS ON IMAGE AND SIGNAL PROCESSING, BIOMEDICAL ENGINEERING AND INFORMATICS (CISP-BMEI 2016), 2016, : 796 - 800