Deepfakes as a threat to a speaker and facial recognition: An overview of tools and attack vectors
Deepfakes present an emerging threat in cyberspace. Recent developments in machine
learning make deepfakes highly believable, and very difficult to differentiate between what is …
learning make deepfakes highly believable, and very difficult to differentiate between what is …
Deepfake Speech Recognition and Detection
HC Chang - International Journal of Pattern Recognition and …, 2023 - World Scientific
Deepfake technology, especially deep voice, which has been derived from artificial
intelligence in recent years, is potentially harmful, and the public is not yet wary. However …
intelligence in recent years, is potentially harmful, and the public is not yet wary. However …
[PDF][PDF] Differential-based Deepfake Speech Detection
BV STANĚK - theses.cz
Deepfake speech technology, which can create highly realistic fake audio, poses significant
challenges, from enabling multi-million dollar scams to complicating legal evidence's …
challenges, from enabling multi-million dollar scams to complicating legal evidence's …
Inconspicuous Audio Anonymization: A Real-Time Approach with a Neural Network Model
Abstract Automatic Speaker Recognition (ASR) systems present an increasingly significant
risk to the privacy of individuals by facilitating the unwitting identification and possible …
risk to the privacy of individuals by facilitating the unwitting identification and possible …