000			01904nam a2200241 4500
008			250513t2019 bx a|||| |||| 00| 0deng d
020			_qhardback
040			_aUniversiti Teknologi Brunei _beng _cUTB
084			_aUTB 120 REPORT THESIS & DISSERTATION _aRTDS 272
100	1		_aManilitphone Thephavanh _eauthor
245	1	0	_aEmotion Analysis / _cManilitphone Thephavanh
260			_aBandar Seri Begawan : _b Universiti Teknologi Brunei , _c©2019
300			_aix, 53 pages : _bcolor illustrations ; _c30 cm
500			_aReport submitted for the degree of BCs in Creative Multimedia Universiti Teknologi Brunei
500			_aAbstract Emotion analysis is useful for many applications dealing with human emotions. A lot of time consuming in order to find the best practice and algorithm for emotion classification, detection and prediction because the analysis can be done in several techniques such as using Ekman's Facial Action Coding System (FACS), Geometric Feature, MPEG 4, facial landmarks detection by Dlib, etc. However, it's hard to achieve high accurate result with certain database and learning tools. This analysis purpose to explore and experiment various of machine learning techniques that can outcome with accuracy results for each technique in a certain condition. By using the Extended Cohn-Kanade Dataset (CK+) with appropriate data mining tool will help to analyze universal of human emotions and result with reasonable accuracy. This analysis focusing on facial landmarks features in order to classify different types of emotion.
504			_aIncludes bibliographical references
610		4	_vReport _aUniversiti Teknologi Brunei
650		4	_aEmotions _xResearch
650		4	_aEmotions _xPsychological aspects
942			_2lc _n0 _cRTDS
998			_eReport, Thesis & Dissertation _s850448 : 002029 c.1_UTB
998			_eCD-ROM _s850449 : CD No. RTDS CD 23 UTB
999			_c23436 _d23436