HOME > 상세정보

상세정보

(A) hybrid machine learning approach towards efficient objective evaluation of facial paralysis based on iris and salient points detection

(A) hybrid machine learning approach towards efficient objective evaluation of facial paralysis based on iris and salient points detection

자료유형
학위논문
개인저자
Barbosa, Jocelyn
서명 / 저자사항
(A) hybrid machine learning approach towards efficient objective evaluation of facial paralysis based on iris and salient points detection / Jocelyn Bilar Barbosa
발행사항
Seoul :   Graduate School, Korea University,   2019  
형태사항
viii, 100장 : 삽화(일부천연색), 도표 ; 26 cm
기타형태 저록
A Hybrid Machine Learning Approach Towards Efficient Objective Evaluation of Facial Paralysis Based on Iris and Salient Points Detection   (DCOLL211009)000000084348  
학위논문주기
학위논문(박사)-- 고려대학교 대학원: 컴퓨터·전파통신공학과, 2019. 8
학과코드
0510   6YD36   361  
일반주기
지도교수: 강재우  
서지주기
참고문헌: 장 89-100
이용가능한 다른형태자료
PDF 파일로도 이용가능;   Requires PDF file reader(application/pdf)  
비통제주제어
Facial paralysis classification, Facial paralysis objective evaluation, Hybrid Classifier, Ensemble of Regression Trees, Iris Detection, Salient Points Detection, Second degree polynomial function, Optimized Daugman's algorithm,,
000 00000nam c2200205 c 4500
001 000045999029
005 20191017124843
007 ta
008 190628s2019 ulkad bmAC 000c eng
040 ▼a 211009 ▼c 211009 ▼d 211009
085 0 ▼a 0510 ▼2 KDCP
090 ▼a 0510 ▼b 6YD36 ▼c 361
100 1 ▼a Barbosa, Jocelyn
245 1 1 ▼a (A) hybrid machine learning approach towards efficient objective evaluation of facial paralysis based on iris and salient points detection / ▼d Jocelyn Bilar Barbosa
260 ▼a Seoul : ▼b Graduate School, Korea University, ▼c 2019
300 ▼a viii, 100장 : ▼b 삽화(일부천연색), 도표 ; ▼c 26 cm
500 ▼a 지도교수: 강재우
502 1 ▼a 학위논문(박사)-- ▼b 고려대학교 대학원: ▼c 컴퓨터·전파통신공학과, ▼d 2019. 8
504 ▼a 참고문헌: 장 89-100
530 ▼a PDF 파일로도 이용가능; ▼c Requires PDF file reader(application/pdf)
653 ▼a Facial paralysis classification, Facial paralysis objective evaluation, Hybrid Classifier, Ensemble of Regression Trees, Iris Detection, Salient Points Detection, Second degree polynomial function, Optimized Daugman's algorithm
776 0 ▼t A Hybrid Machine Learning Approach Towards Efficient Objective Evaluation of Facial Paralysis Based on Iris and Salient Points Detection ▼w (DCOLL211009)000000084348
900 1 0 ▼a 강재우 ▼g 姜在雨, ▼e 지도교수
945 ▼a KLPA

전자정보

No. 원문명 서비스
1
(A) hybrid machine learning approach towards efficient objective evaluation of facial paralysis based on iris and salient points detection (18회 열람)
PDF 초록 목차
No. 소장처 청구기호 등록번호 도서상태 반납예정일 예약 서비스
No. 1 소장처 과학도서관/학위논문서고/ 청구기호 0510 6YD36 361 등록번호 123062317 도서상태 대출가능 반납예정일 예약 서비스 B M
No. 2 소장처 과학도서관/학위논문서고/ 청구기호 0510 6YD36 361 등록번호 123062318 도서상태 대출가능 반납예정일 예약 서비스 B M
No. 3 소장처 세종학술정보원/5층 학위논문실/ 청구기호 0510 6YD36 361 등록번호 153083341 도서상태 대출가능 반납예정일 예약 서비스
No. 소장처 청구기호 등록번호 도서상태 반납예정일 예약 서비스
No. 1 소장처 과학도서관/학위논문서고/ 청구기호 0510 6YD36 361 등록번호 123062317 도서상태 대출가능 반납예정일 예약 서비스 B M
No. 2 소장처 과학도서관/학위논문서고/ 청구기호 0510 6YD36 361 등록번호 123062318 도서상태 대출가능 반납예정일 예약 서비스 B M
No. 소장처 청구기호 등록번호 도서상태 반납예정일 예약 서비스
No. 1 소장처 세종학술정보원/5층 학위논문실/ 청구기호 0510 6YD36 361 등록번호 153083341 도서상태 대출가능 반납예정일 예약 서비스

컨텐츠정보

초록

Facial palsy or paralysis (FP) is a symptom or neuromotor dysfunction that losses voluntary muscles movement in one side of the human face, which could be very devastating in the part of the patients. As the face is the basic means of social interactions and emotional expressions among humans, individuals afflicted can often be introverted and may develop psychological distress, which can be even more severe than the physical disability. This study addresses the problem of objective facial paralysis classification and evaluation. We mainly conducted two scientific researches, which aim to develop an efficient objective facial paralysis classification and grading, utilizing the diagnosis of the neurologist or physicians as the ground truth of the study.

In the first study, we introduce a novel approach for a quantitative assessment of facial paralysis that tackles classification problem for FP type and degree of severity. Specifically, a novel method of quantitative assessment is presented: an algorithm that extracts the human iris and detects facial landmarks; and a hybrid approach combining the rule-based and machine learning algorithm to analyze and prognosticate facial paralysis using the captured images. A method combining the optimized Daugman’s algorithm and Localized Active Contour (LAC) model is proposed to efficiently extract the iris and facial landmark or key points. To improve the performance of LAC, appropriate parameters of initial evolving curve for facial features’ segmentation are automatically selected. The symmetry score is measured by the ratio between features extracted from the two sides of the face. Hybrid classifiers (i.e. rule-based with regularized logistic regression) were employed for discriminating healthy and unhealthy subjects, FP type classification, and for facial paralysis grading based on House-Brackmann (H-B) scale. Quantitative analysis was performed to evaluate the performance of the proposed approach. Experiments show that the proposed method demonstrates its efficiency.

In the second study, we present another method that advances our previous approach for objective FP evaluation and classification, which is crucial for deciding the medical treatment scheme. In particular, we proposed a method based on the ensemble of regression trees to further improve the efficiency of extracting facial salient points and detecting iris or sclera boundaries, particularly with peculiar facial image inputs (e.g. with eyeglasses, excessive beard, eyelid occlusion, etc.). We also employ $2^{nd}$ degree polynomial of parabolic function to improve Daugman's algorithm for detecting occluded iris boundaries, thereby allowing us to efficiently get the area of the iris. Similar to our previous work, the symmetry score of each face is measured by calculating the ratio of both iris area and the distances between the key points in both sides of the face. We build a model by employing hybrid classifier that discriminates healthy from unhealthy subjects and performs FP classification. Objective analysis was conducted to evaluate the performance of the proposed method. As we explore the effect of data augmentation using publicly available datasets of facial expressions, experiments reveal that the proposed approach improves efficiency of the previous method.

목차

1 Introduction 1
 1.1 Facial Paralysis and Diagnosis 1
  1.1.1 Facial Paralysis 1
  1.1.2 Facial Paralysis Assessment Methods in Clinical Practices 4
  1.1.3 Computer-aided Facial Paralysis Diagnosis 5
 1.2 Objectives of the Thesis 7
 1.3 Overview of the Thesis 8
2 Related Prior Work 9
 2.1 Facial Image Acquisition Methods for Computer Aided Facial Paralysis Diagnosis 9
  2.1.1 Image-based Data Acquisition Approach 11
  2.1.2 Video-based Acquisition Approach 12
 2.2 Variations of Methods for Facial Features Extraction and FP Assessment 13
  2.2.1 Face Normalization 13
  2.2.2 Facial Feature Segmentation 13
  2.2.3 Deformation Extraction 14
  2.2.4 Motion Extraction 19
 2.3 Overview of Localized Region-based Active Contour Model (LACM)  25
 2.4 Daugmans Algorithm 27
 2.5 Histogram of Oriented Gradients(HOG) 28
 2.6 Ensemble of Regression Trees 29
 2.7 Shape Invariant Split Tests 30
 2.8 Regularized Logistic Regression 31
3 Efficient quantitative assessment of facial paralysis using iris segmentation and active contour-based key points detection with hybrid classifier 33
 3.1 Background and Problem Definition 33
 3.2 Methods 35
  3.2.1 Proposed Facial Paralysis Assessment: An Overview 35
  3.2.2 Feature Extraction with Optimized Daugmans Integro-Differential 
Operator and Localized Active Contour 37
  3.2.3 Feature Extraction Process 38
  3.2.4 Key points detection 39
  3.2.5 Iris detection 43
  3.2.6 Symmetry Measurement by Iris and Key points 46
 3.3 Results and Discussion 52
  3.3.1 Facial Palsy Classification and Quantitative Assessment of Overall 
Paralysis 54
  3.3.2 Discussion 60
4 paraFaceTest: an Ensemble of Regression Tree-based Facial Features 
Extraction for Efficient Facial Paralysis Classification 62
 4.1 Background and Problem Definition 62
 4.2 Methods 64
  4.2.1 Facial Feature Extraction 66
  4.2.2 Salient Points Detection with Ensemble of Regression Tree 66
  4.2.3 Iris Detection with Optimized Daugmans Algorithm using 2nd Degree Polynomial Function 68
  4.2.4 Facial Paralysis Measurement 72
  4.2.5 Facial Paralysis Classification and Grading 75
 4.3 Results and Discussion 77
  4.3.1 Facial Palsy Classification and Quantitative Assessment of Overall 
Paralysis 78
5 Conclusion 87
Bibliography 89