| 논문명 | 딥러닝기반 실내이미지 디자인스타일 자동분류/Deep-learning based Auto-classifying Design Style of Interior Image/3-학술발표1 |
|---|---|
| 저자명 | 김진성(Kim, Jinsung) ; 김하얀(Kim, Hayan) ; 이진국(Lee, Jin-Kook) |
| 발행사 | 한국실내디자인학회 |
| 수록사항 | 한국실내디자인학회 학술발표대회 논문집 , 제19권 2호 |
| 페이지 | 시작페이지(95) 총페이지(4) |
| 주제분류 | 계획및설계 |
| 주제어 | 실내디자인, 이미지 인식, 딥러닝, 실내디자인 스타일, 자동분류 Interior Design, Image Recognition, Deep Learning, Interior Design Style, Auto-classification |
| 요약2 | This research aims to develop the deep-learning based method for auto-classifying the design style of interior image. The increasing incoming and interest to the interior design caused the trend of self-interior and reasonable consumption of interior style-related products. In the field of interior design, image data is much effective and typical communication tool than text-based document to convey related information. The conventional approach to gain interior related information from the interior image generally depend on the experts’ knowledge or additional research. This approach needs much time for retrieving required information through the plenty of image database. Therefore, the precise and effective process for retrieving interior design specific information is required for saving time and increasing accuracy. The main approach of this research utilizes deep-learning based image recognition technologies to train recognition model and demonstration. This paper describes following steps. 1) Classification of representative interior design style from precedence researches 2) Collecting training interior design style image 3) Training image recognition model 4) Demonstrating trained image recognition model by new test image. Through this research, the possibility of utilizing deep-learning based technologies in the field of interior design was confirmed. |
| 소장처 | 한국실내디자인학회 |