[논문]딥러닝 기반 민화 장르 분류 모델 연구

윤수림; 이영숙

doi:10.9717/kmms.2022.25.10.1524

딥러닝 기반 민화 장르 분류 모델 연구
A Study on the Classification Model of Minhwa Genre Based on Deep Learning 원문보기

멀티미디어학회논문지 = Journal of Korea Multimedia Society, v.25 no.10, 2022년, pp.1524 - 1534

윤수림 (Dept. of Information and Communication Engineering, Dongguk University) , 이영숙 (Institute of Image and Cultural Contents, Dongguk University)

Abstract ▼ AI-Helper

This study proposes the classification model of Minhwa genre based on object detection of deep learning. To detect unique Korean traditional objects in Minhwa, we construct custom datasets by labeling images using object keywords in Minhwa DB. We train YOLOv5 models with custom datasets, and classify images using predicted object labels result, the output of model training. The algorithm consists of two classification steps: 1) according to the painting technique and 2) genre of Minhwa. Through classifying paintings using this algorithm on the Internet, it is expected that the correct information of Minhwa can be built and provided to users forward.

주제어

표/그림 (10)

표 Table 1. Test result of pretrained YOLOv5 with Hojakdo.
그림 Fig, 1. Process of Minhwa classification algorithm.
표 Table 2. Labels of Minhwa, Korean painting, Picture.
표 Table 3. Labeled images of Hojakdo (Minhwa), Korean painting and Picture.
표 Table 4. The object labels of Minhwa.
표 Table 5. Labeled images of Minhwa.
표 Table 6. mAP-50 of YOLOv5 model training with Minhwa, Korean painting, Picture.
표 Table 7. Predicted results of YOLOv5 model training with Hojakdo, Korean painting, Picture
표 Table 8. mAP-50 results of YOLOv5 model training with Minhwa.
표 Table 9. Predicted results of YOLOv5 model training (Minhwa).

참고문헌 (12)

Korean paintings(2022), https://trends.google.co.kr/trends/?geoKR (accessed August 01, 2022).
National Museum of Modern and Contemporary Art(MMCA) VR Art Korean wave Video Series Popularity(2022), http://www.artenews.co.kr/news/articleView.html?idxno784 (accessed August 01, 2022).
E. Shin, "The Study on the Establishment of Monk-Tablet-Style Portraits(Weepae-hyeong Jinyeong) and Folk Features in Late Joseon Dynasty," Journal of Korean Folk P ainting Association, No. 4, pp. 18-39, 2013.12.
E. Choi and Y. Lee, "Designing a Classification System for Minhwa DB," Journal of Korea Multimedia Society, Vol. 25, No. 1, pp. 135-143, 2022.
S. Yang and Y. Lee, "A Study of an AI Based Content Source Data Generation Model using Folk Paintings and Genre Paintings," Journal of Korea Multimedia Society, Vol. 24, No. 5, pp. 736-743, 2020.
I. Kim, M. Lee, and Y. Jeon, "Comparative Analysis of Defect Detection Using YOLO of Deep Learning," Journal of the Korean Society of Manufacturing Technology Engineers, Vol. 30, No. 6, pp. 514-519, 2021
J. Ryu and D. Kwak, "Flame Detection Based on Deep Learning Using HSV Color Model and Forner Detection Algorithm," Fire Science and Engineering, Vol. 35, No. 2, pp. 108-114, 2021.
J. Park and Y. Kim, "A Study on Deep Learning Performance Improvement Based on YOLOv5," The Journal of Korean Institute of Communications and Information Sciences, pp. 1,592-1,593, 2022.
S. Jeon, J. Lee, M. Kim, S. Hong, J. Bang, and W. Kim, "A Study on Helmet Wear Identification Using YOLOv5 Object Recognition Model," The Journal of Korean Institute of Communications and Information Sciences, pp. 1,293-1,294, 2022.
YOLOv5, https://pytorch.org/hub/ultralytics_yolov5/ (accessed August 02, 2022).
B. Kim, "Trend of Image Classification Technology Based on Deep Learning," The Journal of Korean Institute of Communications and Information Sciences, Vol. 35, No. 12, pp. 8-14, 2018.
S. Jeon, S. Muhammad, J. Lee, and M. Jeong, "Road Network Traffic Data Prediction Using ResNet," Journal of Korean Society for Geospatial Information Science, Vol. 29, No. 2, pp. 75-84, 2022.

내보내기 구분	파일저장 인쇄 메일전송
구성항목	기본정보 상세정보 관리번호, 논문명, 저널/프로시딩명, 저자 , 발행년, 권, 호, 시작페이지, 끝페이지, 발행기관 관리번호, 논문명, 대등논문명, 저자 , 저널/프로시딩명, 발행기관, 발행년, 발행언어, 권, 호, 시작페이지, 끝페이지, ISBN, ISSN, 주제분야, 키워드, 초록(한글), 초록(영문), 저자(소속기관)
저장형식	Text(ASCII format) Excel format RefWorks Direct Export RIS format (for Reference Manager, ProCite, EndNote), Scholar's Aids, Mendeley
메일정보	받는사람 (필수) @ 보내는사람 (선택) @ 제목 내용 KISTI 검색결과 이메일 서비스
안내	총 건의 자료가 검색되었습니다. 다운받으실 자료의 인덱스를 입력하세요. (1-10,000) 검색결과의 순서대로 최대 10,000건 까지 다운로드가 가능합니다. 데이타가 많을 경우 속도가 느려질 수 있습니다.(최대 2~3분 소요) 다운로드 파일은 UTF-8 형태로 저장됩니다. 파일의 내용이 제대로 보이지 않을실 때는 웹브라우저 상단의 보기 -> 인코딩 -> 자동선택 여부를 확인하십시오. ~ Text(ASCII format) Excel format

연합인증