Indoor-Outdoor Classification of Audio-Video and Detection of Tampered Audio

Abstract: An improved method for indoor-outdoor classification of audio and video is presented. In second part it explains the detection of tampered audio. For indoor-outdoor classification of audio RT60 approach is used. This method calculates the reverberation time and decay curve of impulse response of the given audio track. The reverberation time value is compared with the threshold value for performing the classification. The contrast of hue component of HSV image of video frame is calculated for video classification. The tampering detection is performed by comparing the reverberation time of original and corrupted audio. The proposed solution proves to be useful in identifying fake audio tracks. It also helps in scene classification in videos ie to detect whether the given video is indoor or outdoor. Keywords: Include RT60, HSV.