Population-wide prediction of severe COVID from electronic healthcare records
15:15 - 15:30
AI Infrastructure as a Platform for Futuristic Applications
15:30 - 16:00
COMPUTER VISION TRACK II
16:00 - 16:30
Turning GANs into Useful Consumer Products
16:30 - 16:50
Layered Neural Atlases for Consistent Video Editing
16:50 - 17:10
Classic Signal Processing meets Deep Learning
17:10 - 17:30
Mads Nielsen is professor of Computer Science at University of Copenhagen, Co-director of the Pioneer Centre in AI, Founder of Cerebriu A/S, Biomediq A/S, Aiomic aps. His major research has been in medical image analysis and its mathematical foundations with a specialty in neurological disorders, breast cancer, osteoporosis, osteoarthritis and lately also in Covid-19. He has obtained more than 20 patents and has published more than 300 papers. In 2012 he co-auhtored with Andrew Ng one of the first papers on deep learning in medical image analysis.
As head of the Technological Infrastructure Division, Aviv Zeevi is responsible for generating collaborations between the industry and academia that create advanced technologies and innovative products; these collaborations strengthen the Israeli industry's long-term technological advantage in the fierce competition of international markets. Zeevi has vast experience in the field of technological developments, in developing online systems as well as training and simulation systems. In his most recent role he served as the head of the ICT department in the Israeli Research Directorate of the European Research Program. Zeevi has a PhD in Information Systems Management from the Tel Aviv University. He holds several additional degrees in Economics, Political Science and Management.
In this talk I will cover how to get a GAN to run on a mobile device and give a consumer control over its results. At Lightricks we build tools for the creative process. These tools need to be consistent, controllable and accessible in order to succeed in the market. GANs are amazing machines with many capabilities, but turning these wild horses into tools used by millions on mobile devices is a challenge. I will describe the stages of turning GANs into product grade features that can even work on the edge. The hurdles of better data, model efficiency and the tradeoffs between control and expression when choosing your architecture.
Ofir Bibi is the VP Research at Lightricks. Bringing the latest and greatest in ML to the core of consumer products, creating efficient research pipelines, methodologies and growing great researchers. His research interests are in the fields of efficient machine learning, statistical signal processing, computational photography and computer graphics. He has taken leading research positions in building systems for estimation and prediction, market optimization and recommendations, but his true passion is in solving challenges by a precise mix of engineering and research.
While image editing and manipulation tools have seen remarkable progress, video editing remains a difficult task that poses two key challenges: (i) edits need to be applied in a temporally consistent manner to all frames, and (ii) editing interfaces need to be able to represent temporal content in an intuitive manner. Thus, video editing has been largely restricted to the domain of professionals. In this talk, I’ll present a new method that tackles these challenges, and allows easy and intuitive editing of everyday videos by novice users.
The pillar of the approach is a novel decomposition of the input video into a set of layered 2D atlases ('texture maps'), each providing a unified representation of an object/background over the entire video. Using the learned decomposition, we can simply edit the 2D atlases (or a single frame), and automatically propagate the edits to the entire video. By operating purely in 2D, our method does not require any prior 3D knowledge about scene geometry or camera poses. I’ll show a variety of exciting editing results including texture mapping, video style transfer, image-to-video texture transfer, and segmentation/labeling propagation.
Project page: https://layered-neural-atlases.github.io/
Tali Dekel is an Assistant Professor at the Mathematics and Computer Science Department at the Weizmann Institute, Israel. She is also a Staff Research Scientist at Google, developing algorithms at the intersection of computer vision, computer graphics, and machine learning. Before Google, she was a Postdoctoral Associate at the Computer Science and Artificial Intelligence Lab (CSAIL) at MIT. Tali completed her Ph.D. studies at the school of electrical engineering, Tel-Aviv University, Israel. Her research interests include computational photography, image/video synthesis, geometry and 3D reconstruction. Her awards and honors include the National Postdoctoral Award for Advancing Women in Science (2014), the Rothschild Postdoctoral Fellowship (2015), the SAMSON - Prime Minister's Researcher Recruitment Prize (2019), Best Paper Honorable Mention in CVPR 2019, and Best Paper Award (Marr Prize) in ICCV 2019.
In this talk, we will show how one may use tools from signal processing to improve neural network performance in various applications including image super-resolution, robustness to label noise, better representation capabilities and more.
Raja Giryes is an associate professor in the school of electrical engineering at Tel Aviv University. His research interests lie at the intersection between signal and image processing and machine learning, and in particular, in deep learning, inverse problems, sparse representations, computational photography, and signal and image modeling. Raja received the EURASIP best P.hD. award, the ERC-StG grant, Maof prize for excellent young faculty (2016-2019), VATAT scholarship for excellent postdoctoral fellows (2014-2015), Intel Research and Excellence Award (2005, 2013), the Excellence in Signal Processing Award (ESPA) from Texas Instruments (2008) and was part of the Azrieli Fellows program (2010-2013). He is an associate editor in IEEE Transactions on Image Processing and Elsevier Pattern Recognition and has organized workshops and tutorials on deep learning theory in various conferences including ICML, CVPR, and ICCV. He serves as a consultant in various high-tech companies including Innoviz technologies and developed a technology that was used as the basis for the MultiVu technologies startup.