Deep neural networks and deep feature selection for spoofing detection in voice authentication systems

• Grant number: 19/21464-1
• Support Opportunities: Scholarships in Brazil - Post-Doctoral
• Start date: August 01, 2023
• End date: November 30, 2024
• Field of knowledge: Engineering - Electrical Engineering
• Principal Investigator: Rodrigo Capobianco Guido
• Grantee: Rodrigo Colnago Contreras
• Host Institution: Instituto de Biociências, Letras e Ciências Exatas (IBILCE). Universidade Estadual Paulista (UNESP). Campus de São José do Rio Preto. São José do Rio Preto , SP, Brazil

Abstract

Voice-based authentication systems have been increasingly used due to their simplicity in relation to the necessary equipment and their versatility, as any smartphone can be used to remotely authenticate an individual. However, this type of system is vulnerable to spoofing attacks, since an impostor can use a recorded voice to be properly authenticated as an authentic user, which is known as a replay attack. Notably, several studies have been conducted to alleviate the problem and thereby increase the reliability of voice authentication systems. Thus, the objective of this project is to improve techniques based on deep learning and deep features to develop solutions that mitigate replay attack fraud. The ASVspoof database, which has been used as a baseline in this international biometric authentication fraud detection challenge, will serve as a basis for evaluating the proposed methodology. Particularly, the intention is to participate in the next edition of the event, which is scheduled for 2023, presenting the research results. In addition, we intend to build a library in Python and C/C++ languages, implementing the material developed in a scalable way for both academia and business environments. Finally, the results obtained should be published in international journals and conferences of recognized excellence.

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

CONTRERAS, RODRIGO COLNAGO; DA SILVA, VITOR TREVELIN XAVIER; DA SILVA, IGOR TREVELIN XAVIER; VIANA, MONIQUE SIMPLICIO; DOS SANTOS, FRANCISCO LLEDO; ZANIN, RODRIGO BRUNO; MARTINS, ERICO FERNANDES OLIVEIRA; GUIDO, RODRIGO CAPOBIANCO. Genetic Algorithm for Feature Selection Applied to Financial Time Series Monotonicity Prediction: Experimental Cases in Cryptocurrencies and Brazilian Assets. Entropy, v. 26, n. 3, p. 22-pg., 2024-03-01. (19/21464-1, 21/12407-4, 22/05186-4, 13/07375-0, 23/06611-3)

VIANA, MONIQUE SIMPLICIO; CONTRERAS, RODRIGO COLNAGO; PESSOA, PAULO CAVALCANTI; DOS SANTOS BONGARTI, MARCELO ADRIANO; ZAMANI, HODA; GUIDO, RODRIGO CAPOBIANCO; MORANDIN JUNIOR, ORIDES. Massive Conscious Neighborhood-Based Crow Search Algorithm for the Pseudo-Coloring Problem. ADVANCES IN SWARM INTELLIGENCE, PT I, ICSI 2024, v. 14788, p. 15-pg., 2024-01-01. (22/05186-4, 19/21464-1, 21/12407-4, 23/06611-3)

CONTRERAS, RODRIGO COLNAGO; HECK, GUSTAVO LUIZ; VIANA, MONIQUE SIMPLICIO; DOS SANTOS BONGARTI, MARCELO ADRIANO; ZAMANI, HODA; GUIDO, RODRIGO CAPOBIANCO. Metaheuristic Algorithms for Enhancing Multicepstral Representation in Voice Spoofing Detection: An Experimental Approach. ADVANCES IN SWARM INTELLIGENCE, PT I, ICSI 2024, v. 14788, p. 16-pg., 2024-01-01. (22/05186-4, 19/21464-1, 21/12407-4, 23/06611-3)