Matt Leotta, Ph.D.
Matt Leotta, Ph.D., is a technical leader on Kitware’s Computer Vision Team located in Clifton Park, New York. The projects he leads primarily focus on 3D reconstruction from imagery and video. However, Matt has also contributed to super-resolution, object detection, and tracking research projects. He has received funding from various government agencies and commercial organizations.
Matt was the principal investigator (PI) on Kitware’s team for the Intelligence Advanced Research Projects Activity (IARPA) CORE3D program. As PI, he led three universities and two outside companies in the collaborative development of Danesfield, an open source framework for 3D semantic reconstruction of buildings from satellite imagery.
Matt is the founder and lead maintainer of TeleSculptor, Kitware’s open source desktop application for 3D reconstruction from aerial video. He is also one of the founders and maintainers of the KWIVER toolkit on which TeleSculptor and other applications were built. This 3D computer vision software is the culmination of work on several Small Business Innovation Research (SBIR) projects for which Matt served as PI or has mentored the PI. The initial work started with an SBIR project with Air Force Research Laboratory (AFRL) in 2013 and has since been extended with additional SBIR funding from AFRL, Army Night Vision and Electronic Sensors Directorate (NVESD), United States Special Operations Command (SOCOM), and the National Geospatial-Intelligence Agency (NGA).
Matt led a commercial effort to develop algorithms for visual navigation of an endoscope for medical applications that resulted in US Patent 10169875.
In addition to his projects, several of Matt’s papers have been published in peer-reviewed international conferences and journals. He also organized and presented tutorials at the Conference on Computer Vision and Pattern Recognition (CVPR) in 2012 and 2015 on the topics of open source computer vision using Python and open source structure-from-motion.
Matt is also involved in Kitware’s Open Source Software Technology Program. He interviews candidates for the program and helps select the computer vision research interns. Matt also mentors students during their internship at Kitware.
Matt received his Ph.D. in computer engineering from Brown University in 2010. Under the supervision of Professor Joseph Mundy, Matt’s work focused on tracking vehicles in traffic videos while simultaneously reconstructing 3D models of the vehicles by fitting a generic deformable model. In 2007, Matt also received his master’s degree in applied mathematics from Brown. He received his bachelor’s degree in computer science and computer systems engineering from Rensselaer Polytechnic Institute in 2003. He graduated summa cum laude. During his graduate and undergraduate studies, Matt worked as a research assistant in robotics and computer vision.
Ph.D. in computer engineering from Brown University, 2010
M.S. in applied mathematics from Brown University, 2007
B.S. in computer science and computer systems engineering, summa cum laude, from Rensselaer Polytechnic Institute, 2003
Best Paper Award presented by EarthVision Workshop, CVPR, 2019
M. Leotta, C. Long, B. Jacquet, M. Zins, D. Lipsa, J. Shan, B. Xu, Z. Li, X. Zhang, S. Chang, M. Purri, J. Xue, and K. Dana, “Urban Semantic 3D Reconstruction From Multiview Satellite Imagery,” in The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops: EarthVision, 2019. [URL]
Get to Know Matt
What is your favorite thing about working at Kitware? My favorite thing about working at Kitware is the opportunity to work on the research problems that I’m interested in and, in doing so, influence the direction of the company. The directions of Kitware’s work are largely driven bottom-up by the employees rather than mandated top-down by senior management. Any employee can get involved with writing proposals to bring in funding for interesting new research and help shape the type of work that Kitware does.
What do you love most about what you do? I love the opportunity to work on challenging, real-world research problems, develop software solutions, and then release that software as open source for anyone to use for free. I love that the work we do to solve a problem for one customer can often help a completely unrelated user that we were not even aware of, and this often leads to new collaborations or new customers.
Share something interesting about yourself that is not on your resume. Outside of work I like to tinker with electronics and design 3D objects for printing on my 3D printer. You can also find me acting as a videographer, editor, and special effects artist for the movies that my pre-teen daughters like to write and star in. Sometimes these hobbies spill over into work life. I have various 3D printed conversation pieces in my office and sometimes the video special effects sneak into video conference calls.
Professional Associations & Service
- Member of the Computer Vision Foundation (CVF)
- Served as corporate relations chair for North America for the International Conference on Pattern Recognition (ICPR) 2020
- Program committee for CVPR
- Program committee for the European Conference on Computer Vision (ECCV)
- Program committee for International Conference on Computer Vision (ICCV)
- Program committee for the IEEE Winter Conference on Applications of Computer Vision (WACV)
- Program committee for the British Machine Vision Conference (BMVC)
- Program committee for the Asian Conference on Computer Vision (ACCV)
- Program committee for the Association for the Advancement of Artificial Intelligence (AAAI)
Matt’s publication list is below. To see all of Kitware’s computer vision publications, please visit the Computer Vision Publications page.
- M. Leotta, J. Shan, X. Zhang, C. Long, B. Xu, M. Purri, M. Zins, B. Jacquet, K. Dana, S. Seida, M. Berlin, Z. Li, J. Xue, and D. Lipsa, "Danesfield: Integrating Deep Learning and Classical Methods for Multiview Semantic 3D Modeling," in Proceedings of the MSS National Symposium on Sensor and Data Fusion, 2019.
- M. Leotta, C. Long, B. Jacquet, M. Zins, D. Lipsa, J. Shan, B. Xu, Z. Li, X. Zhang, S. Chang, M. Purri, J. Xue, and K. Dana, "Urban Semantic 3D Reconstruction From Multiview Satellite Imagery," in The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops: EarthVision, 2019. Winner, Best Paper Award. [URL]
- M. Leotta, E. Smith, and D. Russell, "TeleSculptor: Dense 3D Models from Uncalibrated FMV," in Proceedings of the MSS National Symposium on Passive Sensors, 2018.
- M. Leotta, E. Smith, M. Dawkins, and P. Tunison, "Open source structure-from-motion for aerial video," in Proceedings of the IEEE Winter Conference on Applications of Computer Vision, 2016. [URL]
- M. Leotta, P. Tunison, E. Smith, and M. Dawkins, "MAP-Tk: Motion imagery Aerial Photogrammetry Toolkit," in Proceedings of the MSS National Symposium on Passive Sensors, 2015.
- K. Fieldhouse, M. Leotta, A. Basharat, R. Blue, D. Stoup, C. Atkins, L. Sherrill, B. Boeckel, P. Tunison, J. Becker, M. Dawkins, M. Woehlke, R. Collins, M. Turek, and A. Hoogs, "KWIVER: An open source cross-platform video exploitation framework," in Proceedings of the IEEE Applied Imagery Pattern Recognition Workshop, 2014. [URL]
- Z. Sun, M. Leotta, A. Hoogs, R. Blue, R. Neuroth, J. Vasquez, A. Perera, M. Turek, and E. Blasch, "Vehicle change detection from aerial imagery using detection response maps," in SPIE Defense, Security, and Sensing Motion Imagery for ISR and Situational Awareness, 2014. [URL]
- M. Leotta and J. Mundy, "Vehicle surveillance with a generic, adaptive, 3D vehicle model," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 33, no. 7, pp. 1457-1469, Jul. 2011. [URL]
- A. Perera, S. Oh, M. Leotta, I. Kim, B. Byun, C. Lee, S. McCloskey, B. Miller, Z. Huang, A. Vahdat, W. Yang, G. Mori, K. Tang, D. Koller, L. Fei-Fei, K. Li, G. Chen, J. Corso, Y. Fu, R. Srihari, Y. Fu, and R. Srihari, "GENIE TRECVID 2011 Multimedia Event Detection : Late-Fusion Approaches to Combine Multiple Audio-Visual features," in NIST TRECVID Workshop, 2011.
- M. Leotta, "Generic, deformable models for 3-d vehicle surveillance," Brown University, 2010.
- M. Leotta and J. Mundy, "Predicting high resolution image edges with a generic, adaptive, 3-D vehicle model," in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, 2009. [URL]
- C. Tsai, B. Madore, M. Leotta, M. Sofka, G. Yang, A. Majerovics, H. Tanenbaum, C. Stewart, and B. Roysam, "Automated retinal image analysis over the internet," IEEE Transactions on Information Technology in Biomedicine, vol. 12, no. 4, pp. 480-487, Jul. 2008. [URL]
- M. Leotta, A. Vandergon, and G. Taubin, "3D slit scanning with planar constraints," Computer Graphics Forum, vol. 27, no. 8, pp. 2066-2080, Dec. 2008. [URL]
- M. Leotta, A. Vandergon, and G. Taubin, "Interactive 3D ScanningWithout Tracking," in Brazilian Symposium on Computer Graphics and Image Processing, 2007. [URL]
- M. Leotta and J. Mundy, "Epipolar curve tracking in 3-D," in Proceedings of the IEEE International Conference on Image Processing, 2007. [URL]
- M. Leotta and J. Mundy, "Learning background and shadow appearance with 3-D vehicle models," in Proceedings of the British Machine Vision Conference, 2006. [URL]
- D. Han, M. Leotta, D. Cooper, and J. Mundy, "Vehicle class recognition from video-based on 3D curve probes," in IEEE International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance, 2005. [URL]