Simon A. J. Winder, PhD MEng
Research Scientist (Research Gate)
Current Research Interests
- Machine learning and Artificial Intelligence
- Computer vision and sensing for small robots
- Reinforcement learning
- Neuroscience of perception and action
- Building embodied intelligent agents
- Information theory applied to fundamental physics
- AI exploration of theories of physics
- Number theory
Prior Experience
- Master's degree in Electrical Engineering with first class honors
School of Electrical Engineering, Bath, UK, 1986-1990 - Electronic Engineer
GEC Marconi Avionics / GEC Sensors, Basildon, UK, 1989-1991 - PhD in Computational Neuroscience and Visual Perception
School of Mathematical Sciences, Bath, UK, 1991-1996 - Software Engineer in developing MPEG-4 video compression
Partnership in Advanced Computing Technologies, Bristol University, SGS-Thompson Microelectronics, Bristol, UK, 1996-1997 - Extensive career at Microsoft / Microsoft Research
Microsoft, Redmond, WA, USA, 1997-2015- Image and video data compression, MPEG-4 standardisation
- Real-time optical-flow based video frame interpolation
- Computer vision and 3D graphics
- Real-time image feature matching and tracking
- Robust model estimation for augmented reality
- 3D reconstruction
- Panoramic image stitching
- GPU programming for real-time multi-view stereo
- Data parallel optimization for signal processing and image manipulation
- Scale-space and wavelet representations, and statistics of natural images and video
- Extracting scale-invariant image descriptors for multi-view matching in large media datasets
- Robust probabilistic data-fitting
- Machine learning of image patch embeddings
- Convolutional neural networks for image understanding
- Research Consultant
Impressive Machines LLC, USA, Murley Moss Research, UK, 2015-
Publications and Patents
During my time at Microsoft Research I co-authored papers and patents with:
- Michael Cohen Senior Fellow, Meta
- Matthew Brown Principal Scientist, Wayve
- Richard Hartley Professor and Group Head, Computer Vision Group, Australian National University
- Gang Hua Vice President, Multimodal Experiences Research Lab, Dolby Laboratories
- Sing Bing Kang Distinguished Scientist, Zillow Group
- Richard Szeliski Distinguished Scientist, Google DeepMind
- Desney Tan Principal Researcher and Director, Microsoft Research Medical Devices Group
- Matt Uyttendaele Formerly at Facebook (Meta)
- Larry Zitnick Lead Research Scientist, Facebook AI Research (Meta)
Below you will find a selection of my publications:
-
Discriminant Learning of Local Image Descriptors
in IEEE Transactions on Pattern Analysis and Machine Intelligence, IEEE Computer Society, Vol.33, No.1, pp. 43-57, January, 2011.
High-Quality Multi-Pass Image Resampling
MSR Tech Report no. MSR-TR-2010-10, February 2010.
Picking the Best Daisy
in Computer Vision and Pattern Recognition, IEEE Computer Society, June 2009.
CueFlik: Interactive Concept Learning in Image Search
in CHI 2008 Conference on Human Factors in Computing Systems, 2008.
Discriminant Embedding for Local Image Descriptors
in International Conference on Computer Vision, October 2007.
Learning Local Image Descriptors
in IEEE Computer Society Conf. on Computer Vision and Pattern Recognition, June 2007.
Object Instance Recognition Using Triplets of Feature Symbols
MSR Tech Report no. MSR-TR-2007-53, April 2007.
Multi-Image Matching using Multi-Scale Oriented Patches
in IEEE Computer Society Conf. on Computer Vision and Pattern Recognition, IEEE Computer Society, June 2005.
Multi-Scale Oriented Patches
MSR Tech Report no. MSR-TR-2004-133, December 2004.
Free-Viewpoint Video with Stereo and Matting
in Picture Coding Symposium, San Francisco, CA, December 2004.
High-Quality Video View Interpolation Using a Layered Representation
in ACM SIGGRAPH, Association for Computing Machinery, Inc., August 2004.
Image-based Interactive Exploration of Real-World Environments
in IEEE Computer Graphics & Applications (CG&A), vol. 24, no. 3, pp. 52-63, IEEE Computer Society, May 2004.
High-quality Image-based Interactive Exploration of Real-World Environments
MSR Tech Report no. MSR-TR-2003-61, October 2003.
Including Contrast Gain Control in a Parallel Simulation of the Visual Cortex
in Computational Neursocience: Trends in Research 1995, J. M. Bower (editor), Academic Press Inc., 297-302, 1996.
Simulation of Retinal Contrast Gain Control Mechanisms
in Computational Neursocience: Trends in Research 1995, J. M. Bower (editor), Academic Press Inc., 291-296, 1996.
Current Developments in MPEG-4 Video
First Advanced Digital Video Compression Engineering Conference, 89-96. Presented at ADVICE'96, Jesus College Cambridge, UK, July 8-11, 1996.
A Model for Biological Winner-Take-All Neural Competition Employing Inhibitory using Modulation of NMDA-Mediated Excitatory Gain
Neurocomputing 26-27, 1999, 587-592.
Computer Simulation of Interactions Responsible for Length and Width Tuning in the Striate Cortex
presented at AVA'96, University of Reading, UK, April 1-3, 1996.-
Image Descriptor Quantization
April 2014, Patent: US8712159B21 -
Surface Normal Computation on Noisy Samples of Points
May 2015, Patent: US9043186B21