Simon Jones

SimonJonesPic
B.Eng., Ph.D.
Email: sim DOT mcjones AT gmail DOT com
Google Scholar

Welcome to my academic webpage - you can find the following sections below.

News
About Me
Research Interests
Publications/Code

News

[2014/05/23] - Overdue update to the "About Me" section of the website
[2014/06/06] - Video spotlight available for Unsupervised Spectral Dual Clustering
[2014/04/17] - Camera ready editions of CVPR papers available below.
[2014/03/11] - The code for my CVPR papers is now linked and updated below. Usage instructions included with both.
[2014/03/08] - My 2 submitted papers have been accepted to CVPR 2014 - 1 oral and 1 poster. See you in Columbus!
[2014/03/07] - Academic webpage created.

About Me

I am currently an R&D Engineer at Oxehealth in Oxford. Prior to that, I was an algorithm developer at Hawk-Eye Innovations, working primarily on visual tracking for tennis and other sports.

From 2010 to 2014, I was a PhD student at The University of Sheffield (2010-2014), studying Human Action Analysis in the Department of Electronic and Electrical Engineering under the supervision of Dr. Ling Shao. The title of my thesis was Unsupervised and Semi-supervised Methods for Human Action Analysis. During my undergraduate years (2004-2008) I studied Artificial Intelligence and Software Engineering at The University of Edinburgh, and did my undergraduate dissertation under the supervision of Dr. Hiroshi Shimodaira.

In between my studies I have worked at Data Connection (now known as Metaswitch Networks) and Barclays Capital (now known as Barclays Investment Bank) during which time I've developed a strong interest in many programming languages, including C/C++, Java, C#, Python, Perl and MATLAB.

Outside of work, I enjoy various hobbies, including learning Chinese, various personal programming projects (currently investigating self-learning robotics using deep learning and reinforcement learning), running and cooking.

Research Interests

During the course of my Ph.D., I have been primarily interested in learning from unlabeled data, and how such techniques can be applied in the context of human action analysis. This has prompted me to explore topics such as information retrieval, localisation, temporal video segmentation, clustering and manifold learning, all of which are presented in my final thesis.

More generally, I am also interested in more general machine learning topics, such as deep learning, and am strongly interested in Artificial General Intelligence (AGI), as exemplified in the OpenCog project.

Publications/Code:

Simon Jones and Ling Shao, "Unsupervised Spectral Dual Assignment Clustering of Human Actions in Context," Accepted for presentation at IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Columbus, USA, 2014
[Oral Presentation - Acceptance Rate: 5.75%] [DAKM code]

Simon Jones and Ling Shao, "A Multigraph Representation for Improved Unsupervised/Semi-supervised Learning of Human Actions," Accepted for presentation at IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Columbus, USA, 2014
[Multigraph code (Works on Linux only!)]

Simon Jones and Ling Shao, "Linear Regression Motion Analysis for Unsupervised Temporal Segmentation of Human Actions," Accepted for presentation at IEEE Winter Conference on Applications of Computer Vision (WACV), Steamboat Springs, USA, 2014

Jun Tang, Ling Shao and Simon Jones, "Point Pattern Matching Based on Line Graph Spectral Context and Descriptor Embedding," Accepted for presentation at IEEE Winter Conference on Applications of Computer Vision (WACV), Steamboat Springs, USA, 2014

Simon Jones, Ling Shao and Kairan Du, "Active Learning for Human Action Retrieval Using Query Pool Selection," Neurocomputing, Vol. 124, pp89--96, 2014

Ling Shao and Simon Jones, "Efficient Search and Localisation of Human Actions in Video Databases," Transactions on Systems and Circuits for Video Technology (TCSVT), vol. 24, no. 3, pp. 504-512, Mar. 2014

Simon Jones and Ling Shao, "Rapid Localisation and Retrieval of Human Actions with Relevance Feedback,'' International Conference on Computer Analysis of Images and Patterns (CAIP), York, UK, 2013
[Oral Presentation]

Simon Jones and Ling Shao, "Content-based Retrieval of Human Actions from Realistic Video Databases," Information Sciences, Vol. 236, pp56--65, 2012.

Simon Jones and Ling Shao, "Action Retrieval with Relevance Feedback on YouTube Videos," Proceedings of International Conference on Internet Multimedia Computing and Service (ICIMCS), Chengdu, China, 2011.

Simon Jones, Ling Shao, Jianguo Zhang, and Yan Liu, "Relevance Feedback for Real-world Human Action Retrieval," Pattern Recognition Letters, Vol. 33, Issue 4, pp446-452, 2011.