Date of Publication :8th May 2024

Abstract:The spread of more complex fake methods has increased the need for strong digital audio identification systems, which can be used in a wide range of real-life situations to both stop fraud and look into it. These kinds of programs are necessary to stop problems like fake proof, copyright violations, and people getting into your data without permission. This study talks about a new kind of automatic identification system that can tell the difference between fake and real sounds. It is based on three basic psychoacoustic principles of how humans perceive sound. The main idea behind our suggested system is to make it work like the human hearing system, which can tell the difference between real and fake audio records. Additionally, our system has the amazing ability to tell the difference between sound caught in different places using the same microphone, which makes it much more useful. To do this, we use psychoacoustic principles to pull out traits from the noise data that are useful. Gaussian mixture models are used to analyze these traits. This makes it easier for computers to make decisions automatically, which authenticates the audio material and sorts the recording surroundings into groups. Our automatic identification system is the best at stopping audio fraud and manipulation because it uses advanced statistical models and principles of how people hear sounds. It can be used in many areas, from forensics to property protection and data security, making it a flexible answer to many of the important problems that the digital age has brought up.

Reference :