Matthew Brown is a Lecturer (Assistant Professor) in at the University of Bath. He holds an MEng in Electrical and Information Sciences (Cambridge, 2000) and a PhD in Computer Science (UBC, 2005). Prior to joining the Media Technology Research Centre at Bath he was a postdoctoral researcher at EPFL (2009-2011), UBC (2008-2009) and Microsoft Research (2006-2007).
His research interests are in Computer Vision, Machine Learning and Environmental Informatics. Recent research projects include learning low-level visual representations, recognition and reconstruction in photo collections, and building energy modelling.
See my Google Scholar Profile
Dr. Matthew Brown,
Dept of Computer Science,
East Building 3.2,
University of Bath, Bath
BA2 7AY, UK
(google map, campus map)
- Postdoc in Computer Vision: Dynamic Scene Understanding we have a Postdoc/RA position available supervised by me as part of the OAK project. This will focus on recognition and segmentation in dynamic scenes. For more information and to apply online please go here.
- PhD Scholarships Every year the University of Bath awards a limited number of scholarships to students with excellent academic potential. If this sounds like you, please submit an application via the University applications portal by December/early January for programmes starting in the next academic year.
- Please see my List of Undergraduate Final Year Projects (requires Bath University login).
- RGB-NIR Imaging for Scene Recognition Silicon's natural NIR sensitivity improves visual scene recognition.
- Building Energy Informatics. Data-driven modelling and prediction of building energy usage.
- Learning Local Image
Descriptors. Learn optimal features from training data. See also our online patch data.
- Minimal Solutions for Panoramic Stitching. 2-point and 3-point algorithms for rotational homographies.
- City-Scale Location Recognition. Image recognition for large geotagged datasets.
- 3D Object Recognition and Reconstruction. Fully automatic structure and motion for unordered image sets.
- Multi-Image Matching using Multi-Scale Oriented Patches (MOPS). Minimalist local feature matching.
- AUTOSTITCH. The first fully automatic 2D image stitcher. Check it out on the iPhone, iPad or Android.
- Interactive Image Segmentation using a GMMRF model. Image matting and compositing using statistical models of texture.
I am CTO of Cloudburst Research Inc. Check out AutoStitch Panorama for iPhone, iPad or Android.
My work on panoramic stitching appears in several commercial products, including Autopano Pro, Calico Panorama and Serif Panorama Plus.
I have also contributed to several Microsoft products including Photosynth and MSR Image Composite Editor.
- Chicago Sun. "iPhone camera apps that will bring out your best photographer". July 14 2011.
- LA Times. "A top five of photo apps for the iPhone". July 13 2011.
- New York Times. AutoStitch iPhone: "Magically analyzes the resulting shots and combines them, beautifully". 27th April 2011.
- USA Today "Get more from your smart phone's camera". July 1 2010
- CNN. Watch Bill Gates demo the Microsoft image stitcher on CNN news here (wmv 14Mb). I worked on this project as an intern at MSR.
- New Scientist. "The Whole Shooting Match". Short article on the Autostitch panorama stitcher. 18th October 2003, page 25.
- Forbes. "Student Develops Software for Digital Panoramas". 2nd October 2003
- See my TRIPS PAGES