Gabor and HOG approach to facial emotion recognition

Ryan Melaugh; NH Siddique; SA Coleman; Pratheepan Yogarajah

Gabor and HOG approach to facial emotion recognition

Department of Employment and Learning (DEL)

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

493 Downloads (Pure)

Abstract

Automated facial emotion recognition is an essential step for proper Human-Machine Interaction (HMI) since much of human-human interaction occurs outside of our speech and tone of voice. While other papers have tried singular approaches to this problem, we explore a combination of Gabor and Histogram of Oriented Gradients (HOG) to accurately recognise emotion from still images. This novel method has out-achieved many competing Gabor alone and HOG alone methods with future work aiming to explore larger databases and classifiers.

Original language	English
Title of host publication	Unknown Host Publication
Publisher	National University of Ireland
Number of pages	8
Publication status	Accepted - 30 Jun 2017
Event	Irish Machine Vision and Image Processing - Maynooth, Ireland Duration: 30 Jun 2017 → …

Conference

Conference	Irish Machine Vision and Image Processing
Period	30/06/17 → …

Keywords

Emotion Recognition
Gabor
HOG
Neural Network
SVM

Access to Document

Paper2017 MelaughAuthor Accepted Manuscript, 380 KB

http://uir.ulster.ac.uk/38413/1/IMVIP_Melaugh.pdf

Cite this

@inproceedings{9458ee0b4f43402fad95da9a524d2bbf,

title = "Gabor and HOG approach to facial emotion recognition",

abstract = "Automated facial emotion recognition is an essential step for proper Human-Machine Interaction (HMI) since much of human-human interaction occurs outside of our speech and tone of voice. While other papers have tried singular approaches to this problem, we explore a combination of Gabor and Histogram of Oriented Gradients (HOG) to accurately recognise emotion from still images. This novel method has out-achieved many competing Gabor alone and HOG alone methods with future work aiming to explore larger databases and classifiers.",

keywords = "Emotion Recognition, Gabor, HOG, Neural Network, SVM",

author = "Ryan Melaugh and NH Siddique and SA Coleman and Pratheepan Yogarajah",

year = "2017",

month = jun,

day = "30",

language = "English",

booktitle = "Unknown Host Publication",

publisher = "National University of Ireland",

address = "Ireland",

note = "Irish Machine Vision and Image Processing ; Conference date: 30-06-2017",

}

TY - GEN

T1 - Gabor and HOG approach to facial emotion recognition

AU - Melaugh, Ryan

AU - Siddique, NH

AU - Coleman, SA

AU - Yogarajah, Pratheepan

PY - 2017/6/30

Y1 - 2017/6/30

N2 - Automated facial emotion recognition is an essential step for proper Human-Machine Interaction (HMI) since much of human-human interaction occurs outside of our speech and tone of voice. While other papers have tried singular approaches to this problem, we explore a combination of Gabor and Histogram of Oriented Gradients (HOG) to accurately recognise emotion from still images. This novel method has out-achieved many competing Gabor alone and HOG alone methods with future work aiming to explore larger databases and classifiers.

AB - Automated facial emotion recognition is an essential step for proper Human-Machine Interaction (HMI) since much of human-human interaction occurs outside of our speech and tone of voice. While other papers have tried singular approaches to this problem, we explore a combination of Gabor and Histogram of Oriented Gradients (HOG) to accurately recognise emotion from still images. This novel method has out-achieved many competing Gabor alone and HOG alone methods with future work aiming to explore larger databases and classifiers.

KW - Emotion Recognition

KW - Gabor

KW - HOG

KW - Neural Network

KW - SVM

M3 - Conference contribution

BT - Unknown Host Publication

PB - National University of Ireland

T2 - Irish Machine Vision and Image Processing

Y2 - 30 June 2017

ER -

Gabor and HOG approach to facial emotion recognition

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this