Christian Leistner
Christian Leistner is a computer-vision researcher whose work spans visual tracking, random-forest learning, and image upscaling. He studied telematics at Graz University of Technology, worked as a research assistant there, and wrote a PhD thesis in the Learning, Recognition & Surveillance group. He held a postdoctoral fellowship at ETH Zurich from 2010 to 2012 and worked in applied research at Microsoft from 2012 to 2016. His 2008 tracking paper received the Koenderink Prize in 2018, and later papers address decision forests and image super-resolution. He has worked in computer vision and machine learning at Amazon Prime Air since April 2016 and has served as managing director of Amazon Development Center Austria GmbH since April 2025.
Timeline
-
2000–2006
Studying telematics
He studies telematics at Graz University of Technology from 2000 to 2006.
-
2006–2010
Research in computer vision
He works as a research assistant at TU Graz from 2006 to 2010. His 2010 PhD thesis in TU Graz’s Learning, Recognition & Surveillance group is titled “Semi-Supervised Ensemble Methods for Computer Vision.”
-
2008
A semi-supervised tracker
At ECCV 2008, he coauthors “Semi-Supervised On-line Boosting for Robust Tracking” with Helmut Grabner and Horst Bischof. The paper applies online learning to adapt classifiers for visual tracking.
-
2010–2016
Postdoctoral work and Microsoft
He is a postdoctoral fellow at ETH Zurich from 2010 to 2012. From 2012 to 2016, he works as a Software Development Engineer and applied researcher at Microsoft; in 2012, he organizes a “Random Forests in Computer Vision” tutorial at DAGM-OAGM in Graz, while affiliated with Microsoft Vexcel Imaging.
-
2013–2015
Decision forests and image upscaling
At CVPR 2013, he coauthors “Alternating Decision Forests,” a method that formulates random-forest training as global loss minimization. At CVPR 2015, he coauthors “Fast and Accurate Image Upscaling With Super-Resolution Forests” with Samuel Schulter and Horst Bischof; the method reconstructs a high-resolution image from a single low-resolution input.
-
2018
A Koenderink Prize
In 2018, the Koenderink Prize goes to his paper “Semi-supervised On-Line Boosting for Robust Tracking,” coauthored with Helmut Grabner and Horst Bischof.
-
Since 2025
Managing director in Austria
He becomes managing director of Amazon Development Center Austria GmbH in April 2025. As of October 2026, he is a research scientist at Amazon Prime Air, working in computer vision and machine learning; he begins that role in April 2016.
Career
- Since 2025Amazon Development Center Austria GmbHManaging Director
- Since 2016Amazon.com, Inc.Research Scientist
- 2012–2016MicrosoftSoftware Development Engineer and Applied Researcher
- 2010–2012ETH ZurichPostdoctoral Fellow
- 2006–2010Graz University of TechnologyResearch Assistant
Education
- 2010Graz University of TechnologyPhD
- 2006Graz University of TechnologyTelematics
Most-cited works
- Semi-supervised On-Line Boosting for Robust TrackingWith Horst Bischof
- Fast and accurate image upscaling with super-resolution forestsWith Horst Bischof
- On-line Random ForestsWith Horst Bischof
- Learning object class detectors from weakly annotated videoWith Cordelia Schmid and Vittorio Ferrari
- PROST: Parallel robust online simple trackingWith Horst Bischof
Sources and credits
Photo: We have no published photo of Christian Leistner yet. Send us a free-licensed photo.
- Dr Christian Leistner - Research Scientist - Amazon Prime Air | XING
- ICG - Learning, Recognition & Surveillance
- Semi-Supervised On-line Boosting for Robust Tracking - Graz University of Technology
- DAGM-OAGM 2012: Tutorials
- CVPR 2013 Open Access Repository
- CVPR 2015 Open Access Repository
- ECVA | European Computer Vision Association
- evi.gv.at | Amazon Development Center Austria GmbH 8020 Graz | Firmenbuch





