Deep Learning in Image Diagnostics: Recent Results and Outlook
Bram van Ginneken, Radboud University Medical Center, Netherlands
Unsupervised Internal Learning
Michal Irani, Weizmann Institute of Science, Israel
Double-identity Biometrics
Davide Maltoni, University of Bologna, Italy
Deep Learning in Image Diagnostics: Recent Results and Outlook
Bram van Ginneken
Radboud University Medical Center
Netherlands
Brief Bio
Bram van Ginneken is Professor of Medical Image Analysis at Radboud University Medical Center. Since 2010, he is co-chair of the Diagnostic Image Analysis Group. He also works for Fraunhofer MEVIS in Bremen, Germany, and is a founder of Thirona, a company that develops software and provides services for medical image analysis. Bram studied Physics at the Eindhoven University of Technology and at Utrecht University. In 2001, he obtained his Ph.D. at the Image Sciences Institute on Computer-Aided Diagnosis in Chest Radiography. He has (co-)authored close to 200 publications in international journals. He is Associate Editor of IEEE Transactions on Medical Imaging and member of the Editorial Board of Medical Image Analysis. He pioneered the concept of challenges in medical image analysis.
Abstract
The field of analyzing medical images with computers started in the early 1960s and for half a century the goal of achieving automated analysis at the level of human experts (radiologists, pathologists, ophthalmologists, etc) remained elusive. This has changed in the last few years with the adoption of deep learning. I will give an overview of the key methodological approaches in image and lesion detection, classification, and segmentation, and illustrate the rapid progress that has been made with various recent results. Next, generative modeling techniques will be introduced and I will discuss how automated analysis systems for medical images can be made more robust with such approaches. In the final part of my lecture, I discuss the implications of these exciting developments for diagnostic medicine in the next decade.
Unsupervised Internal Learning
Michal Irani
Distinguished IAPR Speaker
Weizmann Institute of Science
Israel
Brief Bio
Michal Irani is a Professor at the Weizmann Institute of Science, Israel, in the Department of Computer Science and Applied Mathematics. She received a B.Sc. degree in Mathematics and Computer Science from the Hebrew University of Jerusalem, and M.Sc. and Ph.D. degrees in Computer Science from the same institution. During 1993-1996 she was a member of the Vision Technologies Laboratory at the Sarnoff Research Center (Princeton). She joined the Weizmann Institute in 1997. Michal's research interests center around computer vision, image processing, and video information analysis. Michal's prizes and honors include the David Sarnoff Research Center Technical Achievement Award (1994), the Yigal Allon three-year Fellowship for Outstanding Young Scientists (1998), the Morris L. Levinson Prize in Mathematics (2003), and the Maria Petrou Prize (awarded by the IAPR) for outstanding contributions to the fields of Computer Vision and Pattern Recognition (2016). She received the ECCV Best Paper Award in 2000 and in 2002, and was awarded the Honorable Mention for the Marr Prize in 2001 and in 2005. In ICCV 2017 Michal received the Helmholtz “Test of Time Award”.
Abstract
In this talk I will show how complex visual inference tasks can be performed in a totally unsupervised way, by exploiting the internal redundancy inside a single visual datum (whether an image or a video). The strong recurrence of information inside a single image/video provides powerful internal examples which suffice for self-supervision, without any prior examples or training data. I will show the power of this approach through a variety of problems, ranging from low-level to high-level vision tasks. I will further show how this approach gives rise to a new paradigm: “Deep Internal Learning”.
Double-identity Biometrics
Davide Maltoni
University of Bologna
Italy
Brief Bio
Davide Maltoni is a Full Professor at University of Bologna (Dept. of Computer Science and Engineering - DISI), and President of DISI unit in Cesena. His research interests are in the area of Pattern Recognition, Computer Vision, Machine Learning and Computational Neuroscience. Davide Maltoni is co-director of the Biometric Systems Laboratory (BioLab), which is internationally known for its research and publications in the field. Several original techniques have been proposed by BioLab team for fingerprint feature extraction, matching and classification, for hand shape verification, for face location and for performance evaluation of biometric systems. Davide Maltoni is co-author of the Handbook of Fingerprint Recognition published by Springer, 2009 and holds three patents on Fingerprint Recognition. He has been elected IAPR (International Association for Pattern Recognition) Fellow 2010.
Abstract
The feasibility of creating double-identity biometrics (faces, fingerprints and iris) poses serious security threats in security applications such as automated border control where biometrics are used to link the identity of a passenger to his/her e-document. For example, if a morphed face, which is similar enough to the biometric features of two subjects, can be included in an e-passport, then two persons can share it. From the pattern recognition point of view creating/detecting double-identity biometrics is particularly intriguing and can help understanding/overcoming some limitations of current state-of-the-art approaches.