High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

A DCNN Approach for Real Time Unconstrained Face Verification

Author(s):

Febi Elsa Aji , St.Thomas College Of Engineering and Technology Chengannur; Anjali Aravind, St.Thomas College Of Engineering and Technology Chengannur; Athul Ramachandran, St.Thomas College Of Engineering and Technology Chengannur; Sanal Samuel, St.Thomas College Of Engineering and Technology Chengannur; Dr. V. Sheeja Kumari, St.Thomas College Of Engineering and Technology Chengannur

Keywords:

Facial Image Representation, Component Based Face Recognition, Texture Features, DCNN

Abstract

Face verification is the task of validating an identity based on the image of a face, and it is vital in each sectors. Automatic identification of explicit image from a video footage and enormous variations in face cannot be recognized in traditional face verification. We propose real-time unconstrained face verification from each image and videos supported deep convolutional neural network features (DCNN) and value it on IARPA Janus Benchmark A (IJB-A) dataset. Our approach consists of both training and testing stages. For training, we tend to perform landmark detection and face detection from the CASIA-WebFace and also the IJB-A datasets to localize and align each face. Next we tend to train our DCNN on the CASIA -WebFace. Results of experimental evaluations on IJB-A is provided.

Other Details

Paper ID: IJSRDV8I50191
Published in: Volume : 8, Issue : 5
Publication Date: 01/08/2020
Page(s): 174-177

Article Preview

Download Article