Intel-link Associative Identity Predictor

Abstract

We present a defined solution using Convolutional Neural network for predicting the probability of association of an individual with others in a gathering. The system combines the convolutional neural network and machine learning algorithms. The integrated applications are face detection, face clustering, and face identifier. This method is capable of rapid classification and provides better performance than the Eigen faces approach. The DLIB algorithm encodes the face into a 128D array and stores it in the file handling system. The system takes video as a source and uses face recognition Chinese whisper clustering algorithm to cluster the faces of an individual. Based on the data from the cluster results, the apriori/association rule mining algorithm finds the associative probability. The system is integrated with google cloud to provide Application-based Interface which makes application programming interface calls