Proceedings of the 3rd International Conference on Integrated Intelligent Computing Communication & Security (ICIIC 2021)

Detection of Partially Occluded Faces Using Convolutional Neural Networks

Authors
HT Chethana, C Nagavi Trisiladevi, M Athreya Shashank
Corresponding Author
HT Chethana
Available Online 13 September 2021.
DOI
https://doi.org/10.2991/ahis.k.210913.010How to use a DOI?
Keywords
Convolutional Neural Networks (CNN), Disguised Faces in the Wild (DFW), Face Landmark Estimation, Image Encodings, Partial Occlusion
Abstract
Partial occlusion in the face refers to the local region of the face with objects such as sunglasses, scarf, hands and beard which leads to loss of information thereby affecting the overall recognition accuracy. It is one of the challenging problems in computer vision. There are many traditional perceptions based models which have become perfect vehicles for identifying partially occluded facial images in unconstrained environments but they fail to be recognized in constrained environments. The images captured under low lighting conditions and noisy situations are called facial images with a constrained environment. The main contribution of this paper is to recognize partially occluded faces using Convolutional Neural Networks (CNN) in a constrained environment. Hence, an attempt is made in this direction to improve the recognition accuracy for partially occluded facial images. Experimental results demonstrated that the proposed system provides a confidence level of 93% and it outperforms the state of art with the other existing partially occluded face recognition algorithms.
Open Access
This is an open access article distributed under the CC BY-NC license.

Download article (PDF)

Cite this article

TY  - CONF
AU  - HT Chethana
AU  - C Nagavi Trisiladevi
AU  - M Athreya Shashank
PY  - 2021
DA  - 2021/09/13
TI  - Detection of Partially Occluded Faces Using Convolutional Neural Networks
BT  - Proceedings of the 3rd International Conference on Integrated Intelligent Computing Communication & Security (ICIIC 2021)
PB  - Atlantis Press
SP  - 69
EP  - 76
SN  - 2589-4900
UR  - https://doi.org/10.2991/ahis.k.210913.010
DO  - https://doi.org/10.2991/ahis.k.210913.010
ID  - Chethana2021
ER  -