This web site is no longer maintained and the content may be outdated.
Please visit for up-to-date information.
No upcoming events...

Home / Graduate / M.S. Theses Completed
  Alime Reyhan Serim, 2003    

Thesis Title

Neural network based face detection using a multi resolution image pyramid


Face detection, which aims to detect the presence and the position of a human face in an image, is the first important step in automated facial image analysis. Face detection has several applications in areas such as biometric identification, content-based image retrieval, video coding, video conferencing, security and intelligent human-computer interfaces. In this work, upright frontal face detection in gray scale images is studied. Our system is based on learning from examples. Training phase of the system includes the construction of appropriate face and nonface samples and the training of neural networks using a bootstrap method. Testing phase of the system includes the testing of an image pyramid with the neural networks using a window scanning technique and the merging of the detection results from different scales of the image pyramid using hierarchical clustering. Dimensionality reduction is applied on training data. One of the networks is trained with the reduced input space. Two merging approaches are developed. Performance of the system is tested on the datasets using different networks and merging approaches.
Boğaziçi University Department of Computer Engineering
Address: 34342 Bebek, Istanbul, TURKEY
Phone: +90 212 359 4523-24 Fax: +90 212 287 2461
general information:   webmaster: