Genuine Selfie detection Algorithm for Social media Using Image Quality Measures

Selfies are the self-speaking photographs that express oneself. Selfies portraits a person's emotion and in turn it became an effective way of expressing oneself. In recent scenario selfies in social media are also considered for authentication enactment. Face recognition is an extensively used authentication technique for security applications. Face recognition security systems go through vulnerabilities such as printed photo, replayed video and 3d mask attacks. Selfie photographs are considered as the most trustful information from social media. Selfies can be forged for untruthful purposes arise security concern in the recent scenario. This paper proposes an anti-spoofing algorithm to detect fake faces from selfies to get rid of spoofing attacks. Image quality measures and local binary pattern are extracted features from data. Naïve Bayes classifier algorithm is employed here to classify data as real or fake. This algorithm is successfully tested with DSI-1 and DSO-1 datasets and exhibit 92.82% accuracy, 93.54% sensitivity, and 92.15% specificity.