Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Jun 7, 2026

Asthma Detection Research Based on Voice Signal Processing and Machine Learning

Asthma Detection Research Based on Voice Signal Processing and Machine Learning

Published on: July 22, 2025

Performance Analysis of Machine Learning Techniques for Voice Spoofing Detection.

Rekha Rani¹, Bal Kishan¹

¹Department of Computer Science & Applications, Maharshi Dayanand University, Rohtak, Haryana, India.

Journal of Voice : Official Journal of the Voice Foundation

|June 5, 2026

Summary

This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Promoting reactant accumulation at solid-liquid interfaces through microstrain-defect engineering in overall water splitting.

Chemical communications (Cambridge, England)·2026

Same author

Core-Shell Engineering of ZIF-67@N-Doped Hollow Carbon Spheres via Diffusion Strategy for Enhanced Oxygen Evolution Reaction.

Chemistry, an Asian journal·2025

Same author

A performance-driven hybrid text-image classification model for multimodal data.

Scientific reports·2025

Same author

Topotactic transformation of zeolitic imidazolate frameworks into high-performance battery type electrodes for supercapattery application.

Dalton transactions (Cambridge, England : 2003)·2024

Same author

Recent advances in the synthesis of antidepressant derivatives: pharmacologic insights for mood disorders.

3 Biotech·2024

Same author

A Comprehensive Review: Pharmacogenomics and Personalized Medicine Customizing Drug Therapy Based on Individual Genetics Profiles.

Zhongguo ying yong sheng li xue za zhi = Zhongguo yingyong shenglixue zazhi = Chinese journal of applied physiology·2024

Same journal

Voice Changes in Laryngeal Intraepithelial Neoplasia and Nonneoplastic Lesions.

Journal of voice : official journal of the Voice Foundation·2026

Same journal

Self-Perception for Vocal Attractiveness and Its Acoustic Correlates Among Non-Dysphonic Adults in Shanghai.

Journal of voice : official journal of the Voice Foundation·2026

Same journal

Endoscopic Laser Thyroarytenoid Myoneurectomy in Botulinum Toxin-Refractory Adductor Spasmodic Dysphonia: A Case Series.

Journal of voice : official journal of the Voice Foundation·2026

Same journal

A Robust Voice-Based Parkinson's Disease Diagnosis Model via Semantic RGB Feature Transformation and Hybrid Deep Learning.

Journal of voice : official journal of the Voice Foundation·2026

Same journal

Immediate Acoustic and Glottographic Effects of Straw Phonation in Air and Water in Children with Repaired Cleft Palate.

Journal of voice : official journal of the Voice Foundation·2026

Same journal

Direction of Cepstral Peak Prominence Change Moderates Divergent Kinematic Responses to Voice Stimulability Testing.

Journal of voice : official journal of the Voice Foundation·2026

See all related articles

This study compares four voice spoofing detection models. ResNet offers good detection with low computational cost, while Wav2Vec 2.0 achieves better error rates but requires more resources.

Area of Science:

Computer Science
Artificial Intelligence
Cybersecurity

Background:

Voice-based authentication is susceptible to spoofing attacks (e.g., text-to-speech, voice conversion), compromising privacy and authenticity.
Existing detection techniques vary in performance based on datasets, computational power, and deployment environments, making model selection challenging.

Purpose of the Study:

To conduct a comparative experimental analysis of four voice spoofing detection models: Support Vector Machine (SVM), Gaussian Mixture Model (GMM), ResNet, and Wav2Vec 2.0.
To evaluate detection effectiveness using Equal Error Rate (EER) and Tandem Detection Cost Function (T-DCF).
To assess computational efficiency via training and inference times.

Main Methods:

Four models (SVM, GMM, ResNet, Wav2Vec 2.0) were tested on the ASVspoof 2019 Logical Access dataset.

Keywords:

Deep models GMM ResNet SVM Wav2Vec 2.0

More Related Videos

Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody

Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody

Published on: September 27, 2024

Related Experiment Videos

Last Updated: Jun 7, 2026

Asthma Detection Research Based on Voice Signal Processing and Machine Learning

Asthma Detection Research Based on Voice Signal Processing and Machine Learning

Published on: July 22, 2025

Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody

Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody

Published on: September 27, 2024

Performance was measured by Equal Error Rate (EER) and Tandem Detection Cost Function (T-DCF).

Computational cost was evaluated through training and inference times.

Main Results:

ResNet demonstrated effective detection with comparatively low computational cost.
Wav2Vec 2.0 achieved good error rates but incurred a significantly higher fine-tuning cost.
GMM and SVM models showed lower computational costs but higher error rates.

Conclusions:

Complex models do not always guarantee superior performance in voice spoofing detection.
The study provides guidance for selecting spoof detection models based on deployment needs, considering both detection performance and runtime analysis, not solely accuracy.