Jamie Shotton

Senior Researcher
in Computer Vision

Microsoft Research
Cambridge, UK
 



 

 

Home

Research

Publications

Presentations

Data

Code

 

Please see my MSR homepage
for up-to-date information.

Research

I am a Researcher at Microsoft Research Cambridge.  Previous to that I worked as a Toshiba Fellow in Japan having completed my PhD at the University of Cambridge, supervised by Andrew Blake and Roberto Cipolla, and

My thesis is entitled Contour and Texture for Visual Recognition of Object Categories.

See below for a video of our new CVPR 2008 prize-winning demo, Real-Time Object Segmentation with Semantic Texton Forests.

Contour for Visual Recognition

We as humans are effortlessly capable of recognising objects from fragments of image contour.  We demonstrated in our ICCV 2005 paper how an automatic system can exploit contour as a powerful cue for image classification and categorical object detection. An improved multi-scale version of this work has been accepted for publication in PAMI.

Example object detection results on the Weizmann horse database.
Green boxes represent correct detections of the horses, red boxes are false positives, and yellow boxes are false negatives.  The fragments of contour used for detection are visualised in the final column.

More contour visualisations.
Our technique was applied to a 17 object class database from TU Graz.  Here are a few examples where the contour fragments used for detection are superimposed.

Texture for Visual Recognition

A second visual cue is texture. Our ECCV 2006 paper proposed TextonBoost for simultaneous automatic object recognition and segmentation, using the repeatable textural properties of objects.  We show how texture, layout, and textural context can be exploited to achieve accurate semantic segmentations of images, as illustrated in the results below and in the videos available here.  An expanded version has been accepted to IJCV.

Example semantic segmentation results.
TextonBoost applied to the MSRC 21 Class Database.

We have recently improved TextonBoost considerably, making it more accurate and much faster.  This work is summarized in our new CVPR 2008 paper, Semantic Texton Forests.  Based on randomized decision forests, our new system is able to run real-time, illustrated in our demo video:

Real-Time Object Segmentation with Semantic Texton Forests
(winner of CVPR 2008 Demo Prize)

Other research interests

Our visual recognition methods have proven useful for semantic photo synthesis.

Our new dense-stereo algorithm can interpolate between different cameras to facilitate eye contact in one-to-one video conferencing.

Other interests include class-specific segmentation, visual robotic navigation, and image search.

This website was published before I joined MSR and is maintained personally for the benefit of the academic community.  Microsoft is in no way associated with or responsible for the content of these legacy pages.

Last updated : 25 October 2012