Face Recognition Bot Using Deep Learning

Abstract

Face recognition has remained as an interesting, ever ending and significant research topic due to its ever-increasing application in various fields. Face recognition bots have drastically decreased human intervention in human identity authentication. Face recognition bots are used in biometric authentication in cooperate offices, video surveillance, content-based data retrieval. And in various social media platforms.in face recognition bots using deep learning there is no human intervention due to which there is no unnecessary delays in processing the data.in this paper a simulated robot system is designed which can be used for fraud prevention, reduce shop lifting activities, criminal detection and reporting. This face recognition bot clicks the snaps of the people and checks the snap with its database to match the faces of the criminals thereby strengthening the security. In this paper, we proposed a multi-face structure for the identification that is extended to the embedded GPUScheme for real-time identification of multiple faces. The paradigm was trained by a network design that minimised the network parameter, and a monitoring technique has been applied to a parameter System to decrease the cycle time while preserving the method appropriate acceptance rate of recognition. Final experimental results showed that this proposed system recognized multiple faces up to 6 at the same time with up to 0.19 seconds of processing time.