Voice Morphing: Two Identities in One Voice
This work addresses a security threat for biometric systems by extending morph attacks to the voice domain, though it is preliminary and incremental compared to existing image-based attacks.
The paper tackles the problem of voice-based biometric security by introducing Voice Identity Morphing (VIM), a method that synthesizes speech samples impersonating two individuals, achieving over 80% success rate at a 1% false match rate on the Librispeech dataset.
In a biometric system, each biometric sample or template is typically associated with a single identity. However, recent research has demonstrated the possibility of generating "morph" biometric samples that can successfully match more than a single identity. Morph attacks are now recognized as a potential security threat to biometric systems. However, most morph attacks have been studied on biometric modalities operating in the image domain, such as face, fingerprint, and iris. In this preliminary work, we introduce Voice Identity Morphing (VIM) - a voice-based morph attack that can synthesize speech samples that impersonate the voice characteristics of a pair of individuals. Our experiments evaluate the vulnerabilities of two popular speaker recognition systems, ECAPA-TDNN and x-vector, to VIM, with a success rate (MMPMR) of over 80% at a false match rate of 1% on the Librispeech dataset.