Contextual Biomedical Image Learning

A biomedical image characterizes rich contextual information that is defined as the interrelationship among shape, appearance, motion, geometry, imaging modality, disease, biological condition, etc. However, most algorithms either ignore such biomedical image context or partially use it; they resort to linear or parametric models and Gaussian noise assumption; and they need manual interaction. In this talk, I will present novel machine learning approaches that leverage biomedical image context for more effective and efficient analysis of biomedical images. In particular, I will address two methods: (i) Shape Regression Machine (SRM) for deformable shape detection and segmentation and (ii) BoostMotion for landmark motion estimation. With no assumption on the data distribution, these nonparametric methods, grounded on a unifying learning framework of boosting, transfer contextual knowledge from expert-annotated database into machine use. Boosting iteratively selects for the given tasks relevant visual features, which are fast to evaluate and hence enable real time processing. I will illustrate the benefits of context learning for biomedical image analysis using real time demonstrations.
S. Kevin Zhou received his PhD. degree in Electrical Engineering from University of Maryland at College Park in 2004. He then joined Siemens Corporate Research, Princeton, New Jersey, as a research scientist and currently he is a project manager. His research interest lies in statistical signal/image processing, computer vision and machine learning, with their applications to biomedical image analysis (especially biomedical image context learning), biometrics and surveillance (especially face and gait recognition), etc. He has written two research monographs: the lead author of the book entitled Unconstrained Face Recognition (with Chellappa and Zhao, Springer) and a coauthor of the book entitled Recognition of Humans and Their Activities Using Video (with Chellappa and Roy-Chowdhury, Morgan & Claypool Publishers), has edited a book on Analysis and Modeling of Faces and Gestures (with Zhao, Tang, and Gong, Springer LNCS), has published over 50 book chapters and peer-reviewed journal and conference papers, and has possessed over 30 provisional and issued patents. He served in the technical program committee of premier computer vision and medical imaging conferences, gave a tutorial talk on Surveillance Biometrics for ICIP 2006, and organized the third international workshop on Analysis and Modeling of Faces and Gestures (AMFG) in conjunction with ICCV 2007. He was identified as Siemens Junior Top Talent in 2006.