콘볼루션 신경망 기반의 안면영상을 이용한 사상체질 분류
Sasang Constitution Classification using Convolutional Neural Network on Facial Images
0
조회
0
다운로드
2
인용
0
참고문헌
키워드
초록
Objectives Sasang constitutional medicine is a traditional Korean medicine that classifies humans into four constitutions in consideration of individual differences in physical, psychological, and physiological characteristics. In this paper, we proposed a method to classify Taeeum person (TE) and Non-Taeeum person (NTE), Soeum person (SE) and Non-Soeum person (NSE), and Soyang person (ST) and Non-Soyang person (NSY) using a convolutional neural network with only facial images. Methods Based on the convolutional neural network VGG16 architecture, transfer learning is carried out on the facial images of 3738 subjects to classify TE and NTE, SE and NSE, and SY and NSY. Data augmentation techniques are used to increase classification performance. Results The classification performance of TE and NTE, SE and NSE, and SY and NSY was 77.24%, 85.17%, and 80.18% by F1 score and 80.02%, 85.96%, and 72.76% by Precision-Recall AUC (Area Under the receiver operating characteristic Curve) respectively. Conclusions It was found that Soeum person had the most heterogeneous facial features as it had the best classification performance compared to the rest of the constitution, followed by Taeeum person and Soyang person. The experimental results showed that there is a possibility to classify constitutions only with facial images. The performance is expected to increase with additional data such as BMI or personality questionnaire.
Abstract
Objectives Sasang constitutional medicine is a traditional Korean medicine that classifies humans into four constitutions in consideration of individual differences in physical, psychological, and physiological characteristics. In this paper, we proposed a method to classify Taeeum person (TE) and Non-Taeeum person (NTE), Soeum person (SE) and Non-Soeum person (NSE), and Soyang person (ST) and Non-Soyang person (NSY) using a convolutional neural network with only facial images. Methods Based on the convolutional neural network VGG16 architecture, transfer learning is carried out on the facial images of 3738 subjects to classify TE and NTE, SE and NSE, and SY and NSY. Data augmentation techniques are used to increase classification performance. Results The classification performance of TE and NTE, SE and NSE, and SY and NSY was 77.24%, 85.17%, and 80.18% by F1 score and 80.02%, 85.96%, and 72.76% by Precision-Recall AUC (Area Under the receiver operating characteristic Curve) respectively. Conclusions It was found that Soeum person had the most heterogeneous facial features as it had the best classification performance compared to the rest of the constitution, followed by Taeeum person and Soyang person. The experimental results showed that there is a possibility to classify constitutions only with facial images. The performance is expected to increase with additional data such as BMI or personality questionnaire.
참고문헌 (22)
- Lee J, "Longevity and Life Preservation in Oriental Medicine" (1996), Kyung НееUniv. Press
- 이혜림, "얼굴영상의 형태학적 특징 추출을 통한 사상체질 분류 시스템", 한국컴퓨터정보학회논문지, 20(8), pp.15~21 (2015), 한국컴퓨터정보학회
- Cho DU, "Sasang Constitution Classification Using Shape Analysis of Face", Proceedings of the Korea Information Processing Society Conference 2006, pp.423~426 (2006), Korea Information Processing Society
- 조동욱, "얼굴의 형태학적 관계 분석에 의한 사상 체질 분류 시스템", 정보처리학회 논문지, 14(3), pp.153~162 (2007), 한국정보처리학회
- 이세환, "영상 및 음성 신호 처리를 이용한 장년기 여성의 사상체질 분류 방법의 제안", 한국산학기술학회논문지, 9(5), pp.1210~1217 (2008), 한국산학기술학회
- 이선영, "안면형상을 활용한 사상체질 진단 연구에 관한 체계적 고찰", 사상체질의학회지, 24(4), pp.17~27 (2012), 사상체질의학회[DOI]
- "Am I a Taeumin? or a Soyangin?… Represent the Standard Facial Features in the Sasang Constitution" (2012)[URL]
- Do JH, "Development of an integrated Sasang constitution diagnosis method using face, body shape, voice, and questionnaire information", BMC complementary and alternative medicine, 12(1), pp.1~9 (2012)[DOI]
- 정지오, "딥 러닝을 이용한 안면 여드름 분류 모델", 한국정보통신학회논문지, 23(4), pp.381~387 (2019), 한국정보통신학회[DOI]
- LeCun Y, "Deep learning", nature, 521(7553), pp.436~444 (2015)[DOI]
- 백영화, "체질정보은행 구축현황에 관한 소고", 사상체질의학회지, 24(2), pp.47~53 (2012), 사상체질의학회[DOI]
- King DE, "Dlib-ml : A machine learning toolkit", The Journal of Machine Learning Research, 10, pp.1755~1758 (2009)
- Gonzalez RC, "Digital image processing" (2008), prentice hall
- D. S. Ma, "The Chicago face database: A free stimulus set of faces and norming data", Behav. Res. Methods, 47, pp.1122~1135 (2015)
- Weiss K, "A survey of transfer learning", Journal of Big data, 3(1), pp.1~40 (2016)[DOI]
- Simonyan K, "Very deep convolutional networks for large-scale image recognition", arXiv preprint arXiv:1409.1556, ~ (2014)
- 유호진, "영상기반 주차공간 분류 딥 모델을 위한 데이터 증강기법", 정보과학회논문지, 49(2), pp.126~136 (2022), 한국정보과학회[DOI]
- 황두성, "클래스 불균형 문제에서베이지안 알고리즘의 학습 행위 분석", 전자공학회논문지 - CI, 45(6), pp.179~186 (2008), 대한전자공학회
- Song HW, "Prediction of dairy cow mastitis with multi-sensor data using Multi-Layer Perceptron (MLP)", Proceedings of the Korea Information Processing Society Conference 2020, pp.788~791 (2020), Korea Information Processing Society
- Pham DD, "Body mass index and facial cues in Sasang typology for young and elderly persons", Evidence-Based Complementary and Alternative Medicine, 2011 (2011)[DOI]
- 장은수, "체형 안면 소증 및 성정 특성 기반 사상체질 진단 프로그램 개발", 사상체질의학회지, 24(1), pp.21~31 (2012), 사상체질의학회
- 석재화, "사상체질진단 두면부 분석프로그램의 Upgrade 연구 :성별？연령별 특징", 사상체질의학회지, 19(3), pp.30~50 (2007), 사상체질의학회