The Effect of FaceAgeing in Face Recogntion

EasyChair Preprint 809

Abstract

Age progression still has many challenges and affects face recognition tasks. Over the past decade, many researchers have been working on face processing system in order to tackle these challenges of face recognition gap especially in the presence of variation in the age of an individual. In this paper, we propose a new preprocessing method that can improve a face for recognition. Here we align the faces into one template and remove a background which can hold some objects that affect the recognition. In particular, in the stage of recognition, we use a convolutional neural network based architecture along with the pre-trained VGG-Face model to extract features. These features are passed into a Cosine similarity classifier classification. A thorough experimental examination of face identification is achieved on the most widely used age database which is FGNET aging with a huge age gap and different ethnicity and gender. Our experimental results on the FGNET face database illustrate that the proposed approach has the ability to outperform the present state of the art methods.

Keyphrases: Age Progression, cosine similarity, face recognition

@booklet{EasyChair:809,
  author    = {Ali Elmahmudi and Hassan Ugail},
  title     = {The Effect of FaceAgeing in Face Recogntion},
  doi       = {10.29007/8vms},
  howpublished = {EasyChair Preprint 809},
  year      = {EasyChair, 2019}}