Speaker: Rui Shen Title: Towards Solving Labeling Problems in Computer Vision Abstract: Many seemingly different computer vision problems can be formulated as labeling problems. A labeling problem deals with assigning a label to a site, such as a pixel in an image, based on the observed information. This talk will focus on solving two such problems using probabilistic methods. The first problem is multi-exposure image fusion, which aims to eliminate under- or over-exposed regions in an image by combining information of the same scene captured under different exposure settings. A generalized random walks framework is proposed to calculate a globally optimal solution subject to two quality measures (local contrast and color consistency) by formulating the fusion problem as probability estimation. The second problem is tuberculosis (TB) cavity detection from chest radiographs (CXRs). TB is a deadly infectious disease and the presence of cavities in the upper lung zones is a strong indicator that the disease has developed into a highly infectious state. Here, we propose an automated TB detection technique from CXRs, which employs a hybrid knowledge-based Bayesian classification using a gradient measure and a circularity measure. Short Bio: Rui Shen received the B.Eng. degree in computer science and technology from Beihang University, Beijing, China in 2005, and the M.S. degree in computing science from the University of Alberta, Edmonton, Canada in 2007. He is a Ph.D. candidate in the department of computing science at the University of Alberta. He is expecting to graduate in April 2012. His current research interests include probability models, sensor fusion, stereo vision, and image processing.