About

My name is Richard Stebbing and I am a software engineer based in the San Francisco Bay Area, specializing in developing and implementing machine learning and computer vision algorithms for a variety of applications. Until earlier this year, I was Director of Machine Learning of a stealth machine intelligence company, and I've most recently been working on applications to support a quantitative hedge fund.

Previously, I completed my DPhil in the Institute of Biomedical Engineering at the University of Oxford where I was funded on a Rhodes Scholarship and supervised by Alison Noble. My DPhil research was on model-based segmentation of 2D and 3D ultrasound images and sequences with a specific focus on echocardiography. During this period, I also did a couple of stints at Microsoft Research where I was supervised by Andrew Fitzgibbon. I am originally from Auckland, New Zealand, where I studied Electrical and Electronic Engineering in the Department of Electrical and Computer Engineering at the University of Auckland. Before this I went to Northcote College.

The primary purpose of this site is to host information and code about projects I have worked on and think may be useful to others. The website is continuously under construction (albeit infrequently updated) so more of these things will become available as I get time to put them up.

Code

B-spline Regression (GitHub)

A repository providing the code necessary to fit explicit uniform B-splines of any degree to unstructured 2D/3D/ND point data.

Subdivision (GitHub)

A repository containing, among other things, C++ and Python code for efficient and exact evaluation of Doo-Sabin subdivision surfaces. This is useful for modelling smooth surfaces exactly using only a small number of control vertices.

Efron's Dice (GitHub / Blog)

A Python program to find non-transitive dice combinations (e.g. Efron's dice).

Shards (GitHub / Blog)

A collection of scripts to build an approximation of an image by greedily optimising overlapping transparent triangles using Levenberg-Marquardt.

Suffix Tree (GitHub)

A pure Python implementation of a suffix tree used to find all (potentially overlapping) repeats in a string.

QPBO Alpha Expansion (GitHub)

A simple C++ implementation (with example Python module) of alpha-expansion moves using QPBO for minimisation of non-submodular energies. Uses Vladimir Kolmogorov's QPBO implementation and is inspired by Shai Bagon's MATLAB wrapper.

Common (GitHub)

General Python and C++ files common to multiple projects.

Powerlifting Meet Manager (GitHub)

A small PyQt4 application written over a couple of days to help manage the Oxford versus Cambridge 2012 varsity meet.

Selected Publications

Data-Driven Shape Parameterization for Segmentation of the Right Ventricle from 3D+t Echocardiography (Article)

R. V. Stebbing, A. I. L. Namburete, R. Upton, P. Leeson, and J. A. Noble. In Medical Image Analysis, 2015.

Learning-Based Prediction of Gestational Age from Ultrasound Images of the Fetal Brain (Article)

A. I. L. Namburete, R. V. Stebbing, B. Kemp, M. Yaqub, A. T. Papageorghiou, J. A. Noble. In Medical Image Analysis, 2015.

User-Specific Hand Modeling from Monocular Depth Sequences (Project / Article)

J. Taylor, R. V. Stebbing, V. Ramakrishna, C. Keskin, J. Shotton, S. Izadi, A. Hertzmann, and A. W. Fitzgibbon. In IEEE CVPR, 2014.

Delineating Anatomical Boundaries Using the Boundary Fragment Model (Article)

R. V. Stebbing and J. A. Noble. In Medical Image Analysis, 2013.

Contact

Email: richie[dot]stebbing[at]gmail[dot]com