Topic: Deep Learning and Deep Vision
Deep learning is hot. Finally, recognition by computer vision works, and computer-aided diagnosis (CAD) successes are reported in many different application fields. The prerequisites are big data, GPU computing and proper network designs, which we now have. But deep learning is also largely a ‘black box’.
In this talk I will first discuss a CAD application with big data: retina screening for early detection of retinal damage due to diabetes, i.e. prevention of blindness. I then will focus on how much the computer vision and the biological vision world can learn from each other. We study the visual system as a geometry inference engine, and deep convolutional neural networks as incremental contextual geometry systems.
Bastian Leibe is a professor at RWTH Aachen University, where he leads the Computer Vision group. He holds an M.Sc. degree from Georgia Institute of Technology (1999), a Diploma degree from the University of Stuttgart (2001) and a PhD from ETH Zurich (2004). After completing his dissertation on visual object categorization at ETH Zurich, he worked as a postdoctoral research associate in the Multimodal Interactive Systems group at TU Darmstadt and in the Computer Vision Laboratory at ETH Zurich. During this time, he developed algorithms for multi-object detection and tracking from mobile platforms, as well as for 3D shape modeling and registration. He has published over 70 articles in peer-reviewed journals and conferences. Over the years, he received several awards for his research work, including the Virtual Reality Best Paper Award in 2000, the ETH Medal and the DAGM Main Prize in 2004, the CVPR Best Paper Award in 2007, the DAGM Olympus Prize in 2008, the ICRA Best Vision Paper Award in 2009 and the ISPRS Journal of Photogrammetry and Remote Sensing Best Paper of the Year 2010 Award.