Cybersecurity · Digital Life
Resources
For the development of this challenge, access to existing solutions will be offered at https://github.com and will work closely with the research team to propose and develop innovative methods.
Are you interested?
If you are a professor or university student and you are interested in participating in the TUTORING program, register your information so that we can start the program.
Subject Area
The Deepfake Speech Detection in the Wild challenge is aimed at students with a strong interest in new technologies, especially those related to artificial intelligence techniques.
User-level Linux proficiency and familiarity with the Python programming language is recommended; as well as an understanding of the fundamentals of machine learning.
Introduction
Today's digital landscape, social media, news or the chatting platforms are susceptible to spoofing attacks, where malicious actors attempt to create fake voices by manipulating speech data.
In the realm of information security, a spoofing attack occurs when an individual or software masquerades as someone else by falsifying data.
In today's digital landscape, social media, news or the chatting platforms are susceptible to spoofing attacks, where malicious actors attempt to create fake voices by manipulating speech data.
There are four primary methods for executing these fake voice attacks: impersonation, voice conversion (VC), text-to-speech (TTS) synthesis, and replay. Impersonation attacks have garnered relatively little attention, as they usually demand specialized expertise, such as that possessed by professional impersonators.
In contrast, VC, TTS, and replay spoofing attacks can all be carried out using readily available software tools and consumer devices, and as a result, they have received more significant attention. The danger posed by these techniques is now widely acknowledged, particularly within academic circles and, increasingly, in the industrial world.
Challenge raised
The challenge will consist in the benchmarking of current state of the art baselines delivered by the Academia, e.g., using technology and research datasets like as in
https://www.asvspoof.org/index2021.html
and the development of deep learning approaches for the real-time detection of fake voices in a telephone channel scenario.
-10-3.jpg "Jordi Luque Serrano")