Abstract: This paper presents a new approach for face recognition of single face,multi faces, and twins faces, based on Convolutional Neural Network (CNN). Analyzing and exploring essential parameters that can influence model performances the approach has been used in this work to recognize face. Furthermore different elegant prior contemporary models are recruited to introduce new leveraging model. Also show how a very large scale dataset (3M images over 3K people) can be assembled. The proposed method has been compared to the most contemporary approaches and conducted on proposed method with LFW, and YTF datasets and it shows an excellent performance with higher accuracy.

Keywords: Deep Face Recognition, Multiscale images.