USER-GENERATED PORNOGRAPHIC VIDEO DETECTION USING SHOT-BASED SENSOR PATTERN NOISE

摘要

Illegal distribution of user-generated pornography (UGP)rnvideos raises lots of negative aspects associated with therndigital contents. This paper proposes a first forensic techniquernto detect UGP video. To discriminate UGP videos,rnwe exploit shot based sensor pattern noise (SPN) from anrnimage under investigation. By thresholding average peakto-rncorrelation (PCE) values from the shot based SPNs, thernvideo is decided as a UGP video. Preliminary experiments,rnwhich consist of UGP videos, normal videos, and Japaneserncommercial pornographic videos, indicate adequate performancernof the proposed technique.