CIF: Small: Nonintrusive Digital Speech Forensics: Source Identification and Content authentication

University Of Maryland, College Park, College Park MD

Investigators

Abstract

Current digital media editing software allows malicious amateur users to perform imperceptible alterations to digital speech communications This creates a serious threat to the "knowledge life cycle". The proposal seeks to develop theories, methods and tools for extracting and visualizing evidence from digital speech content for the purpose of media source identification and content authentication. The strategy will be based on the important paradigm of nonintrusive media forensics. The hypothesis is that the physical devices and associated signal processing chains may leave behind intrinsic "fingerprint" that are detectable by statistical methods. The hypothesis clearly indicates the purpose of the project.