Share Email Print

Spie Press Book

Machine Learning for Face, Emotion, and Pain Recognition
Author(s): Gholamreza Anbarjafari; Pejman Rasti; Fatemeh Noroozi; Jelena Gorbova; Rain Eric Haamer
Format Member Price Non-Member Price

Book Description

This Spotlight explains how to build an automated system for face, emotion, and pain recognition. The steps involved include pre-processing, face detection and segmentation, feature extraction, and finally recognition to classify features and show the accuracy of the system. State-of-the-art algorithms are used to describe all possible solutions of each step. For face detection and segmentation, several approaches are described to detect a face in images: Viola-Jones, color-based approaches, histogram-based approaches, and morphological operation. Local binary patterns, edge detectors, wavelets, discrete cosine transformation, Gabor filters, and fuzzified features are used for feature extraction. The last step includes three approaches for recognition: classification techniques (with a special focus on deep learning), statistical modeling, and distance/similarity measures.

Book Details

Date Published: 30 May 2018
Pages: 106
ISBN: 9781510619876
Volume: SL37

Table of Contents
SHOW Table of Contents | HIDE Table of Contents

1 Introduction

2 Face Recognition
2.1 Introduction
     2.1.1 History of face recognition
     2.1.2 Contributions
     2.1.3 Overview
2.2 Conventional and state-of-the-art techniques in face recognition
     2.2.1 Correlation
     2.2.2 Principal component analysis
     2.2.3 Linear space
     2.2.4 Linear discriminant analysis
     2.2.5 Independent component analysis
     2.2.6 Other conventional methods in face recognition
     2.2.7 Non-negative matrix factorization
     2.2.8 Incremental non-negative matrix factorization
2.3 PDF-based face recognition
     2.3.1 Previous face-recognition system featuring a PDF
     2.3.2 Chromatic-histogram-based human face tracking
     2.3.3 Local binary pattern for face recognition
     2.3.4 Histogram-based face recognition
     2.3.5 PDF-based face recognition
2.4 Data fusion
     2.4.1 Reasons for using ensemble-based systems
     2.4.2 History of ensemble systems
     2.4.3 Converting KLD values into probability values
     2.4.4 Majority voting
     2.4.5 Sum and mean rule
     2.4.6 Weighted sum rule
     2.4.7 Minimum/maximum/median rule
     2.4.8 Product rule
     2.4.9 Feature vector fusion

3 Emotion Recognition
3.1 Speech-based emotion recognition
     3.1.1 Speech-based emotion-recognition elements
3.2 Other modalities
     3.2.1 Text-based emotion recognition
     3.2.2 EEG-based emotion recognition
3.3 Gesture-based emotion recognition
     3.3.1 Affect detection from body gestures
     3.3.2 General structure of body-gesture-recognition methods
     3.3.3 Feature extraction
     3.3.4 Classification methods
3.4 Facial expression recognition
     3.4.1 Measurement of facial expression
     3.4.2 Feature extraction for facial expression analysis
     3.4.3 Classification
     3.4.4 3-D facial expression recognition

4 Pain Recognition

5 Face Databases
5.1 Elicitation methods
     5.1.1 Posed
     5.1.2 Induced
     5.1.3 Spontaneous
5.2 Categories of emotion
5.3 Database types
     5.3.1 Static databases
     5.3.2 Video databases
     5.3.3 Miscellaneous databases
5.4 Pain databases


Automatic facial analysis is a fundamental subject in affective computing. Its main applications involve human-computer interaction. The systems developed for this purpose consider combinations of different modalities, based on vocal and visual cues. This Spotlight takes the foregoing modalities into account to review the conventional and state-of-the-art automatic facial expression analysis systems and their applications. In addition, this ebook surveys the available public databases in this field. We hope to provide readers with a good starting point to start their research in the field of human behavior analysis and facial expression analyses.

Gholamreza Anbarjafari
Jelena Gorbova
Rain Eric Hammer
Pejman Rasti
Fatemeh Noroozi
May 2018

© SPIE. Terms of Use
Back to Top