Endorsement to audio recorded in different acoustic environment with feature as reverberation time with blind reverberation time estimation method

摘要

Recorded Audio location authentication is required when we want to clarify any evidence as recording of an event; audio recording rely on the different aspect like structure, composition, geometry of the audio recording and also a recording device. To authenticate the recording we need to consider many features of event happened. Audio sample remove background noise using spectral subtraction method, second extract the reverberation time of audio sample using tail selection blind reverberation time estimation method. Previous method for authentication of audio sample Sabin and Erying approach for geometric construction of acoustic environment, Electrical network frequency Effective against cut and paste attack, Time Domain analysis based Effective to find editing and compression attacks. Whereas Blind Reverberation time estimation gives effective classification of acoustic environment, as reverberation time depends on acoustic environment for classification Euclidean distance gives well performance of classification. Blind reverberation calculate reverberation time using tail selection approach as tail of audio sample gives more accurate reverberation time.