본문 바로 가기

로고

국내 최대 기계 및 로봇 연구정보
통합검색 화살표
  • Dimension 1200es (3D프린터)
  • 국내학술지

    187 0
    국내학술지 제목 게시판 내용
    제목(국문) 컨볼루셔널 신경망과 케스케이드 안면 특징점 검출기를 이용한 얼굴의 특징점 분류
    제목(영문) Facial Point Classifier using Convolution Neural Network and Cascade Facial Point Detector
    저자 유제훈 (Je-Hun Yu ,중앙대학교 전기전자공학과 ) ▷공저자네트워크등록하기
    고광은 (Kwang-Eun Ko ,중앙대학교 전자전기공학부 ) ▷공저자네트워크등록하기
    심귀보 (Kwee-Bo Sim ,중앙대학교 전기전자공학부 ) ▷공저자네트워크등록하기
    초록
    초록(영문)

    Nowadays many people have an interest in facial expression and the behavior of people. These are human-robot interaction (HRI) researchers utilize digital image processing, pattern recognition and machine learning for their studies. Facial feature point detector algorithms are very important for face recognition, gaze tracking, expression, and emotion recognition. In this paper, a cascade facial feature point detector is used for finding facial feature points such as the eyes, nose and mouth. However, the detector has difficulty extracting the feature points from several images, because images have different conditions such as size, color, brightness, etc. Therefore, in this paper, we propose an algorithm using a modified cascade facial feature point detector using a convolutional neural network. The structure of the convolution neural network is based on LeNet-5 of Yann LeCun. For input data of the convolutional neural network, outputs from a cascade facial feature point detector that have color and gray images were used. The images were resized to 32x32 . In addition, the gray images were made into the TIN format. The gray and color images are the basis for the convolution neural network. Then, we classified about 1,200 testing images that show SUbjects. This research found that the proposed method is more accurate than a cascade facial feature point detector, because the algorithm provides modified results from the cascade facial feature point detector.

    keyword cascade facial point detector, convolutional neural network, TIN format, human-robot interaction
    저널명 제어로봇시스템학회논문지 ▷관련저널보기
    VOL 22
    PAGE 0241
    발표년도 2016
    국문File 국문다운로드
    영문File
    • 페이스북아이콘
    • 트위터 아이콘

    서브 사이드

    서브 우측상단1