Audio Authentication refers to the process of detecting if the audio is real voice or fake/synthesized voice. It is also the process of detecting if the audio has been tampered with or modified.
Now a days, there is a new technology being used for human-like voice synthesis using deep neural network techniques. It is difficult to distinguish it from real human voice by a listening test. This is
referred to as deep fake voice. A culprit is likely to tamper or modify the audio using readily available tools to destroy the proofs. The spectrum clarity is examined and number of other parameters are extracted to detect deep fake voice. The number of tests are carried out on audio to detect phase or spectral changes.

​

Applications:

• AAA5000 contains automatic audio authentication system.​It has the important utility of detecting deep fake audio. The automated algorithm examines the spectrogram before after noise cancellation for audio as a first check and then extracts further parameters to confirm if the voice is real voice of deep fake/synthesized voice using text to speech tool.

• It uses number of tests to check if DC shift, LPC discontinuity, LPC error signal discontinuity, zeros discontinuity, copy-paste detection, Fs detection, audio info, pause/spectrum discontinuity, ENF variation, background variation is detected. The complete process is automated and report is also generated in just 5 minutes.

• It performs seamlessly with great accuracy. It contains bench mark algorithms for deep fake as well as tamper detection.

• There is a facility for all types of basic audio processing like record, play, export, split stereo channels, zoomin, zoomout, audio processing like normalize, show clipping with opening and closing of windows and changing their layout etc.

• All above features are supported for stereo files too.

• Ability to carry out individual tests is also provided.

• Facility to run on the device as a complete offline solution.​