Title: Recognizing Altered Faces Due to Aging and Disguises
Speaker: Richa Singh
Affiliation Ph. D. candidate, West Virginia University
Date: August 23, 2008
Abstract
Automatic face recognition is a long standing problem in computer vision that requires the ability to identify an individual despite several variations in the appearance of face. Current face recognition systems capture faces of cooperative individuals in a controlled environment as part of the face recognition process. It is therefore possible to control lighting, pose, background, and quality of images. However, in a real world application, we have to deal with both ideal and imperfect data. Many applications require face images to be captured outdoors where the lighting conditions are unpredictable, the subjects may not be cooperative, the poses may vary, or the angles and distances from the camera may not be normal. Performance of current face recognition systems significantly deteriorates for such imperfect and challenging cases.
In general, challenges in automatic face recognition can be classified into five categories: illumination, image quality, expression, pose, aging, and disguise. In this presentation, I will be presenting my research on face verification with age progression and variations in disguise. To address these two challenges, first an age transformation algorithm is proposed that registers two face images and minimizes the aging variations. Unlike the conventional method, the gallery face image is transformed with respect to the probe face image and facial features are extracted from the registered gallery and probe face images. Further, a granular approach is proposed for face verification by extracting dynamic feed-forward neural architecture
based 2D log polar Gabor phase features at different granular levels. The granular levels exhibit non-disjoint spatiotemporal features which are combined using the proposed likelihood ratio based Support Vector Machine match score
fusion algorithm. The proposed face verification algorithm is validated using five face
databases. The Notre Dame face database is used for performance evaluation since
it contains comprehensive variations in expression and illumination, and has been
widely used for evaluating face recognition algorithms. In addition, a face
database which contains images with age variations and three face databases specifically aimed
at validating the performance for disguised face images are used for evaluation. The
performance of the proposed face verification algorithm is also compared with
existing face verification algorithms.
Biography
Richa Singh received the M.S. degree in computer science in 2005 and is currently a Ph.D. candidate in computer science at West Virginia University, Morgantown, USA. She had been actively involved in the development of
a multimodal biometric system, which includes face, fingerprint, signature, and
iris recognition at the Indian Institute of Technology, Kanpur, from July 2002 to July
2004. Her current areasof interest include biometrics, pattern recognition, image
processing, machine learning, granular computing, and data fusion. She has 73 publications in
refereed journals, book chapters, and conferences. Richa is a member of the IEEE,
Computer Society and ACM. She is also a member of the Phi Kappa Phi, Tau Beta Pi, Upsilon Pi
Epsilon, and Eta Kappa Nu honor societies. She was the recipient of four best paper awards.