• Home
•  
• Call for Papers
• Paper Submission
• Registration
• Housing
•  
• Student Awards
• Technical Program
• Paper Search
• Plenaries
• Tutorials
• Special Sessions
• Meeting and Social Event Schedule
•  
• Venue
• Social Activities
• Exhibits
• Supporters
•  
• Organizing Committee
• Technical Program Committee
• Reviewers

Technical Program

Paper Detail

Paper:	MA-P6.9
Session:	Image and Video Multiresolution Processing
Time:	Monday, September 17, 09:50 - 12:30
Presentation:	Poster
Title:	AN ADAPTIVE MULTIRESOLUTION APPROACH TO FINGERPRINT RECOGNITION
Authors:	Amina Chebira; Carnegie Mellon University
	Luis P. Coelho; Carnegie Mellon University / University of Pittsburgh
	Aliaksei Sandryhaila; Carnegie Mellon University
	Stephen Lin; Carnegie Mellon University
	William G. Jenkinson; Carnegie Mellon University
	Jeremiah MacSleyne; Carnegie Mellon University
	Christopher Hoffman; Carnegie Mellon University
	Philipp Cuadra; Carnegie Mellon University
	Charles Jackson; Carnegie Mellon University
	Markus Puschel; Carnegie Mellon University
	Jelena Kovacevic; Carnegie Mellon University
Abstract:	We propose an adaptive multiresolution (MR) approach to the classification of fingerprint images. The system adds MR decomposition in front of a generic classifier consisting of feature computation and classification in eachMR subspace, yielding local decisions, which are then combined into a global decision using a weighting algorithm. In our previous work on classification of protein subcellular location images, we showed that the space-frequency localized information in the MR subspaces adds significantly to the discriminative power of the system. Here, we go one step farther; We develop a new weighting method which allows for the discriminative power of each subband to be expressed and examined within each class. This, in turn, allows us to evaluate the importance of the information contained within a specific subband. Moreover, we develop a pruning procedure to eliminate the subbands that do not contain useful information. This leads to potential identification of the appropriate MR decomposition both on a per class basis and for a given dataset. With this new approach, we make the system adaptive, flexible as well as more accurate and efficient.