닫기
Loading..

전자정보연구정보센터 ICT 융합 전문연구정보의 집대성

학술대회 프로시딩

홈 홈 > 연구문헌 > 학술대회 프로시딩 > 한국정보과학회 학술대회 > KSC 2020

KSC 2020

Current Result Document : 9,194 / 9,194

한글제목(Korean Title) Door Handle Classification via Convolutional Neural Network Model
영문제목(English Title) Door Handle Classification via Convolutional Neural Network Model
저자(Author) Andrew Lee  
원문수록처(Citation) VOL 47 NO. 02 PP. 1623 ~ 1625 (2020. 12)
한글내용
(Korean Abstract)
영문내용
(English Abstract)
While robots perform various tasks, there are many obstacles to overcome. One of them is opening a door. Robots utilize different strategies to open doors, depending on the doorknob type. The robot chooses the right door opening procedure if the doorknob type is well recognized, so solving the doorknob classification problem is vital for overcoming the doorknob obstacle. Convolutional neural network (CNN) is a type of deep learning technique that is good for image classification. In this study, a CNN architecture to build a door handle classification model was investigated, and three optimized models with good performances in recall were proposed
키워드(Keyword)
원문 PDF 다운로드