Fall 2017Semester 1Research MethodsFacial Expression of EmotionRecognition: A Survey Submission From
Javed Anjum Sheikh Department of Computer Science andInformation Technology Mind Map Contents Chapter 1. 3 1.1 Introduction: 3 Chapter 2.
4 2.1 Introduction: 4 2.2 Problems: 4 2.3 Conclusion: 5 Chapter 3.
6 3.1 Introduction: 6 3.2 Solutions: 6 3.2.4. Solution of problem 2.2.4.
6 3.3 Conclusion: 7 Chapter 11.1 Introduction:In every field in this world computer plays some role indifferent ways. It helps human to solve their difficult problems in an instantif some knows how to use this machine named computer. Understanding computer isalso a problem and to avoid this problem computer scientist use develop thistheir programs in simplest way so, that user can use it easily.
For thispurpose they introduced a field Human Computer Interaction (HCI) which issolely based on the interaction of human with computer.HCI has many sub-section to dealt with like developingcomputer interface, User Experience, usability etc. Among these sub section thereis section which based on human emotion interaction with machine. This sectionis combination part of Computer Vision and Machine Learning.Whenwe talk to each other, our facial expression explains about our feeling and howwe feels about something. Our words doesn’t deliver our emotions to anotherperson about anything, for this purpose, our emotions participates a lot.Facial expression helps us in different ways like pain detection for patients,surveillance of suspicious person and driver monitoring system which helps toreduce the rate of accidents.Inthis paper we discuss about the types of emotions and their recognition throughdifferent technique and how it helps us in real life world.
This paper dividedinto three parts, part 1 is about the problems that was face by differentauthors in their papers. Parts 2 explains their solution and finally in part 3we conclude the paper.Chapter 22.1 Introduction: Thischapter discusses the problems that were addressed by different author in theirrespective papers to detect the expression of emotions. Next section of thisChapter describes the problem statements against the paper titles, which westudied during our review of different research papers. 2.2 Problems: Sr no Title Problem Statement 2.
2.1 “Face Expression Detection on Kinect using Active Appearance Model and Fuzzy Logic.” The aim of this research is to detect facial expression by observing the change of key features in AAM using Fuzzy Logic. Fuzzy Logic is used to determine the current emotions based on prior knowledge derived from Facial Action Coding System (FACS). 2.2.2 “Human Facial Detection Form Detected in Captured Image using Back Propagation Neural Network” Human expression can be detected by observing the movement of eyes, mouth, nose and lips. Face plays an important role to detect expressions of emotion.
2.2.3 “Facial Expression Recognition” Extraction of appearance based facial features and recognition of four facial expression. 2.2.4 “A Vision-based Facial Expression Recognition and Adaptation System from Video Stream” Develop a real time vision based facial expression recognition and adaptation system for human computer interaction.
Detect, identify and recognize facial expression of user from facial image in real time and to be able to adapt with new user’s facial expression. 2.2.5 “Application of Improved AAM and Probabilistic Neural Network to Facial Expression Recognition.
” Using Improved Active Appearance Model and Probabilistic Neural Network to recognize six facial expression from still images. 2.2.
6 Facial expression analysis with facial expression deformation. Deforming the subtle emotion into extreme emotion and recognize it. 2.
2.7 2.2.8 2.2.
9 2.2.10 2.3 Conclusion: Chapter 33.1 Introduction:This chapter explains thesolutions of the different paper that were discussed in the previous section 3.2 Solutions: 3.2.
1. Solutionof problem 2.2.1 Datasetgenerated by some volunteers having Action Units (AUs) 0 to 5 of FACS wererecorded on Kinect with the feature were extracted by AAM.
The outliers werethen removed by Modified Thompson Tau test. Outliers are defined as data pointsthat are statistically inconsistent with the rest of data. Furthermore, each ofthe AUs were checked using significant test-t sample two different variants (Welch’st-test). The significant AUs then became the input of fuzzy system.
3.2.4. Solution of problem 2.
2.4 Thesolution of problem is illustrated in the below fig. face was detected by atemplate matching approach. The selected face was then converted into agreyscale image and histogram equalization is applied. Eigen vector of the faceimage is then calculated and then a face library formed. A clustering algorithmis used to categorized the face, given to the system, after get trained.
Thissystem had the ability to accept the unknown expression and categorized separately. 3.3 Conclusion: