A fast and robust face detection based on module switching network

  • Authors:
  • Jung-Bae Kim;Young Hoon Sung;Seok-Cheol Kee

  • Affiliations:
  • Human Computer Interaction Lab., Samsung Advanced Institute of Technology, Yongin-si, Gyeonggi-do, Korea;Human Computer Interaction Lab., Samsung Advanced Institute of Technology, Yongin-si, Gyeonggi-do, Korea;Human Computer Interaction Lab., Samsung Advanced Institute of Technology, Yongin-si, Gyeonggi-do, Korea

  • Venue:
  • FGR' 04 Proceedings of the Sixth IEEE international conference on Automatic face and gesture recognition
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

We suggest a face detection method for video surveillance system to store a user's images fast and robustly. Our system consists of face and motion detectors. The face detector adopts a gentle AdaBoost algorithm to detect a face. Employing a module switching network, we extend the detectable facial pose range without loss of time. The motion detector, using temporal edges and temporal variance, decides whether a user exists when the face detector fails to detect a face. The proposed method, implemented in the form of software and a PCI interface card, displays 92.3% detection ratio in the test for CMU test DB.