Jan Kautz
Director of Visual Computing Research | NVIDIA

I lead the Visual Computing Research Group at NVIDIA, working predominantly on computer vision problems — from low-level vision (denoising, super-resolution, computational photography) and geometric vision (structure from motion, SLAM, optical flow) to high-level vision (detection, recognition, classification). Furthermore, NVIDIA is starting a Machine Learning Research Group, which I am currently managing. I am an Affiliate Professor of Visual Computing in the Department of Computer Science at University College London. I have three doctoral students, Dan Calian, Daniyar Tur­mu­kham­be­tov, and Yotam Doron (sponsored by The Foundry), who will be finishing up soon.

News
june
2016
Some more news articles on our work with DARPA on Virtual Eye have appeared: Engadget, Digital Trends, Daily Mail, NVIDIA's blog.
may
2016
Some news articles on our work with DARPA on Virtual Eye have started to appear: Yahoo, Tech Insider, Business Insider, C4ISR & Networks, Tech.Mic, China Topix.
may
2016
I am speaking at the LDV Vision Summit 2016 in New York City. The program is very interesting, come join us.
march
2016
We have two papers accepted at CVPR 2016 — one on gesture recognition and one on representing point clouds with GMMs.
october
2015
We presented our Maximum Likelihood Mixture Decoupling technique at 3DV (oral presentation). It enables very fast and accurate point cloud registration.
june
2015
Our paper on hand gesture recognition using 3D convolutional neural networks won the VIVA hand gesture recognition challenge.
june
2015
Our paper on accelerated screen-space ray tracing was accepted at High-Per­for­mance Graphics 2015.
may
2015
Our paper on physically-based rendering for mixed reality was accepted at the Eurographics Symposium on Rendering 2015.
may
2015
My team at NVIDIA Research is continuing to hire in computational photography and computer vision. In addition, we are now hiring in machine learning. See here.
april
2015
april
2015
Our paper on slim near eye displays using pinholes was published in Applied Optics.
march
2015
Our paper on modeling object appearance through context-conditioned component analysis was accepted to CVPR 2015.
march
2015
Our Local Laplacian Filtering method is a research highlight in this month's Communi­cations of the ACM and has made it to the cover page.
october
2014
NVIDIA Research is hiring in computational photography / optics, computer vision, and computational sensing / UI. See here for more details.
october
2014
I am now leading the Mobile Visual Computing Research Group at NVIDIA Research.
august
2014
Research
Low-Level Vision
We work on a variety of low-level vision problems, from image processing problems, such as denoising, demosaicking and super-resolution. Furthermore, we pursue research in computational photography.
Geometric Vision
My team and I are investigating novel approaches for simultaneous localization and mapping (SLAM), efficient 3D reconstruction methods, 3D processing, as well as stereo and optical flow.
High-Level Vision
Detection, recognition, and classification are important aspects of our work. In particular, we have focused on gesture recognition and video classification using machine learning approaches.
Machine Learning
Recently, we started research into machine learning and are building up a group in this area. We mainly explore deep learning, including convolutional neural networks, recurrent neural networks, and deep reinforcement learning.
Publications
2016
Multilayer and Multimodal Fusion of Deep Neural Networks for Video Classification
X. Yang, P. Molchanov, and J. Kautz
ACM Multimedia
October 2016, to appear, oral
Loss Functions for Neural Networks for Image Processing
H. Zhao, O. Gallo, I. Frosio, J. Kautz
arXiv
June 2016
Accelerated Generative Models for 3D Point Cloud Data
B. Eckart, K. Kim, A. Troccoli, A. Kelly, J. Kautz
IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
June 2016, pages 5497-5505, spotlight oral
Online Detection and Classification of Dynamic Hand Gestures with Recurrent 3D Convolutional Neural Networks
P. Molchanov, X. Yang, S. Gupta, K. Kim, S. Tyree, J. Kautz
IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
June 2016, pages 4207-4215
Towards Selecting Robust Hand Gestures for Automotive Interfaces
S. Gupta, P. Molchanov, X. Yang, K. Kim, S. Tyree, J. Kautz
IEEE Intelligent Vehicles Symposium 2016
June 2016
2015
Robust Model-based 3D Head Pose Estimation
G. P. Meyer, S. Gupta, I. Frosio, D. Reddy, J. Kautz
IEEE International Conference on Computer Vision (ICCV)
December 2015, pages 3649-3657
Interactive Sketch-Driven Image Synthesis
D. Turmukhambetov, N. Campbell, D. Goldman, J. Kautz
Computer Graphics Forum
34(8), December 2015, pages 130-142
Joint 5D Pen Input for Light Field Displays
J. Tompkin, S. Muff, J. McCann, H. Pfister, J. Kautz, M. Alexa, W. Matusik
ACM User Interface Software and Technology (UIST)
November 2015
MLMD: Maximum Likelihood Mixture Decoupling for Fast and Accurate Point Cloud Registration
B. Eckart, K. Kim, A. Troccoli, A. Kelly, J. Kautz
International Conference on 3D Vision
October 2015, pages 241-249
An Adaptive Acceleration Structure for Screen-space Ray Tracing
S. Widmer, D. Pajak, A. Schulz, K. Pulli, J. Kautz, M. Goesele, D. Luebke
High-Performance Graphics 2015
August 2015, pages 67-76
Filtering Environment Illumination for Interactive Physically-Based Rendering in Mixed Reality
S. Mehta, K. Kim, D. Pajak, K. Pulli, J. Kautz, R. Ramamoorthi
Eurographics Symposium on Rendering 2015
June 2015
Modeling Object Appearance using Context-Conditioned Component Analysis
D. Turmukhambetov, N. Campbell, S. Prince, J. Kautz
IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
June 2015, pages 4156-4164
Hand Gesture Recognition with 3D Convolutional Neural Networks
P. Molchanov, S. Gupta, K. Kim, J. Kautz
IEEE CVPR Workshop on Observing and Understanding Hands in Action
June 2015, winner of the VIVA hand gesture challenge
Locally Non-rigid Registration for Mobile HDR Photography
O. Gallo, A. Troccoli, J. Hu, K. Pulli, J. Kautz
IEEE CVPR Embedded Vision Workshop
June 2015
Slim Near Eye Display Using Pinhole Aperture Arrays
K. Aksit, J. Kautz, D. Luebke
Applied Optics
54(11), April 2015, pages 3422-3427
Local Laplacian Filters: Edge-aware Image Processing with a Laplacian Pyramid
S. Paris, S. Hasinoff, J. Kautz
Communications of the ACM – Research Highlight
58(3), March 2015, pages 81-91
Speaker-Following Video Subtitles
Y. Hu, J. Kautz, Y. Yu, and W. Wang
ACM Transactions on Multimedia Computing, Communications, and Applications
11(2), January 2015, pages 32:1-32:17
CV
10/2015 — now
Director of Visual Computing Research, NVIDIA, USA
10/2014 — 09/2015
Senior Research Manager, NVIDIA, USA
Leading the mobile visual computing research group at NVIDIA.
09/2013 — 09/2014
Senior Research Scientist, NVIDIA, USA
Research in comp. photography and computer vision.
10/2015 — ongoing
Affiliate Professor of Visual Computing, University College London, UK
10/2012 — 09/2015
Professor of Visual Computing, University College London, UK
10/2011 — 09/2012
Associate Professor (Reader), University College London, UK
10/2009 — 09/2011
Associate Professor (Senior Lecturer), University College London, UK
03/2006 — 09/2009
Assistant Professor (Lecturer), University College London, UK
Research in visual computing, teaching and supervision of students (BSc, MSc, PhD).
07/2003 — 01/2006
Post-Doctoral Researcher, Massachusetts Institute of Technology, USA
Working on appearance editing and realistic, real-time rendering.
09/1999 — 04/2003
PhD Student, Max-Planck-Institut fur Informatik, Germany
Received PhD (summa cum laude).
05/1998 — 09/1999
Graduate Student, University of Waterloo, Canada
Received MMath.
10/1993 — 01/1999
Student, University Erlangen-Nurnberg, Germany
Received Diplom-Informatiker (MSc in Computer Science).
Contact
Jan Kautz
NVIDIA | Mobile Visual Computing Research Group
2 Technology Park Drive | Westford, MA 01886