Statistical Approaches for Passive Multimedia Forensics

Twenty first century is no doubt an expectancy of digital world. Each individual is drowned into the sea of digital multimedia. Knowingly or unknowingly people come across thousands of pictures/videos in their daily life routines. These contents get instilled somewhere in the human mind. Images and videos are powerful channel to convey a wealth of information as well as emotional impact. Therefore, response of viewer depends on what he/she perceives from the color, texture, shape and moving images or a video. Taking advantage of this fact, these major features are sometimes twisted by some cheat professionals to leave a long lasting emotional impression on the viewers though later on the truth prevails but when harm has already been done. This results in serious consequences, increasing apathy and creating fallacy in many daily life applications and interferences. With the easy availability of video/image capturing hardware and multimedia editing software, it has now become easier to manipulate any digital multimedia contents. Now a days, these kind of manipulations do not require expert hands to use these softwares, in fact it can be used even by novice users. Since digital multimedia contents can be altered, their authenticity cannot be taken for granted.

Horridness of fake digital multimedia contents has already promoted various researchers to evolve some algorithms which can check the authenticity of doubtful contents. Some major contributions have already been seen in image forgery detection, although a lot effort is needed for video forgery detection.