Development of a face recognition system using deep convolutional neural network in a multi-view vision environment

College

Gokongwei College of Engineering

Department/Unit

Electronics And Communications Engg

Document Type

Article

Source Title

International Journal of Advanced Trends in Computer Science and Engineering

Volume

8

Issue

3

First Page

369

Last Page

374

Publication Date

5-1-2019

Abstract

This paper presents the implementation of Face Recognition System for multi-view vision system consisting of three cameras. It captures input frames from RLC423 camera; the main structure composed of recognizing faces, embeddings computation and database vector comparison. OpenCV, Python dlib, face recognition or Openface and ResNet34 are utilized for this face recognition method. It illustrates the application of face recognition and compare it to different dataset with static and dynamic approaches such as MIT-CBCL face recognition dataset, Caltech faces and EPFL-RLC Multi-Camera Dataset which, obtained 1.5% training loss and 90.3% model accuracy. © 2019, World Academy of Research in Science and Engineering. All rights reserved.

html

Digitial Object Identifier (DOI)

10.30534/ijatcse/2019/06832019

Disciplines

Electrical and Computer Engineering

Keywords

Human face recognition (Computer science)

Upload File

wf_yes

This document is currently not available here.

Share

COinS