Pathological Speech Classification Using a Convolutional Neural Network

Nam H. Trinh; Darragh O'Brien

doi:10.21427/9dnc-n002

Pathological Speech Classification Using a Convolutional Neural Network

Nam H. Trinh, Darragh O'Brien

Research output: Contribution to conference › Paper › peer-review

Abstract

Convolutional Neural Networks (CNNs) have enabled significant improvements across a number of applications in computer vision such as object detection, face recognition and image classification. An audio signal can be visually represented as a spectrogram that captures the time-varying frequency content of the signal. This paper describes how a CNN can be applied to the spectrogram of an audio signal to distinguish pathological from healthy speech. We propose a CNN structure and implement it using Keras to test the approach. A classification accuracy of over 95% is obtained in experiments on two public pathological speech datasets.

Original language	English
DOIs	https://doi.org/10.21427/9dnc-n002
Publication status	Published - 1 Jan 2019
Externally published	Yes
Event	IMVIP 2019: Irish Machine Vision & Image Processing - Technological University Dublin, Dublin, Ireland Duration: 28 Aug 2019 → 30 Aug 2019

Conference

Conference	IMVIP 2019: Irish Machine Vision & Image Processing
Country/Territory	Ireland
City	Dublin
Period	28/08/19 → 30/08/19

Keywords

Convolutional Neural Networks
CNNs
computer vision
object detection
face recognition
image classification
spectrogram
audio signal
pathological speech
healthy speech
Keras
classification accuracy

Access to Document

10.21427/9dnc-n002Licence: CC BY-NC-SA

Cite this

@conference{792eae5777084827b09390db94ed2321,

title = "Pathological Speech Classification Using a Convolutional Neural Network",

abstract = "Convolutional Neural Networks (CNNs) have enabled significant improvements across a number of applications in computer vision such as object detection, face recognition and image classification. An audio signal can be visually represented as a spectrogram that captures the time-varying frequency content of the signal. This paper describes how a CNN can be applied to the spectrogram of an audio signal to distinguish pathological from healthy speech. We propose a CNN structure and implement it using Keras to test the approach. A classification accuracy of over 95% is obtained in experiments on two public pathological speech datasets.",

keywords = "Convolutional Neural Networks, CNNs, computer vision, object detection, face recognition, image classification, spectrogram, audio signal, pathological speech, healthy speech, Keras, classification accuracy",

author = "Trinh, {Nam H.} and Darragh O'Brien",

year = "2019",

month = jan,

day = "1",

doi = "10.21427/9dnc-n002",

language = "English",

note = "IMVIP 2019: Irish Machine Vision & Image Processing ; Conference date: 28-08-2019 Through 30-08-2019",

}

TY - CONF

T1 - Pathological Speech Classification Using a Convolutional Neural Network

AU - Trinh, Nam H.

AU - O'Brien, Darragh

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Convolutional Neural Networks (CNNs) have enabled significant improvements across a number of applications in computer vision such as object detection, face recognition and image classification. An audio signal can be visually represented as a spectrogram that captures the time-varying frequency content of the signal. This paper describes how a CNN can be applied to the spectrogram of an audio signal to distinguish pathological from healthy speech. We propose a CNN structure and implement it using Keras to test the approach. A classification accuracy of over 95% is obtained in experiments on two public pathological speech datasets.

AB - Convolutional Neural Networks (CNNs) have enabled significant improvements across a number of applications in computer vision such as object detection, face recognition and image classification. An audio signal can be visually represented as a spectrogram that captures the time-varying frequency content of the signal. This paper describes how a CNN can be applied to the spectrogram of an audio signal to distinguish pathological from healthy speech. We propose a CNN structure and implement it using Keras to test the approach. A classification accuracy of over 95% is obtained in experiments on two public pathological speech datasets.

KW - Convolutional Neural Networks

KW - CNNs

KW - computer vision

KW - object detection

KW - face recognition

KW - image classification

KW - spectrogram

KW - audio signal

KW - pathological speech

KW - healthy speech

KW - Keras

KW - classification accuracy

U2 - 10.21427/9dnc-n002

DO - 10.21427/9dnc-n002

M3 - Paper

T2 - IMVIP 2019: Irish Machine Vision & Image Processing

Y2 - 28 August 2019 through 30 August 2019

ER -

Pathological Speech Classification Using a Convolutional Neural Network

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this