I teach a number of undergraduate and postgraduate masters modules in computer science as outlined below. I also supervise projects for third and forth year undergraduates and postgraduate masters students; scroll down to see a list of the sorts of projects I supervise.
I am a strong proponent of research-led teaching and where possible tie in material to current research. For modules that are focussed on software development, I ensure that current tools and frameworks are used.
The quality of my teaching has been recognised by awards including a Vice-Chancellor's Teaching Award and a SUSU Excellence in Teaching Award. I am a Fellow of the Higher Education Academy which demonstrates my personal commitment to professionalism in learning and teaching in higher education.
I lead teach our differntiable prgramming/deep learning module. I've had a hand in most parts of this module, including developing lectures and labs on everything from automatic differentiation to convolutional networks to generative models, and seeting the coursework assignments.
Together with Adam Prügel-Bennet, I'm part of the lecturing team for the advanced machine learning module. My lectures cover large-scale machine learning. I also help supervise the group projects undertaken by the students.
I've co-taught this module with Mark Nixon since its inception, alongside COMP3204. In 2020-21 I'm the module leader, but the teaching will be done by my colleagues Hansung Kim and Xiaohao Cai.
I teach this module together with Mark Nixon. Mark covers lower-level vision and image processing, whereas I cover higher-level vision topics. My part of the course uses an innovative teaching format that makes use of interactive slides written in OpenIMAJ that demonstrate the workings of core techniques and allow for real world vision experiments to be performed in the lecture theatre. Outside of the lectures, the students have full access to the source code for the slides which they can use to aid learning and help with the coursework assignments. In addition to lecturing, I'm responsible for developing the coursework for this module.
In 2014 I was the Southampton University Students Union faculty winner of an Excellence in Teaching Award in the innovative teaching category for my work on this module.
I supervise third year projects aligned with my research activities. In particular in the areas of multimedia information retrieval; image, video and text mining and classification; big-data; and the synergy between textual data and image data on the web. Projects that contribute to or build upon our OpenIMAJ software are particularly welcome.
Examples of projects that I've supervised include: Internet Meme Spread - A Twitter Image Macro Analysis; A Hybrid Music Recommendation System and Advanced Music Search; Skydiving formation Recognition; Prediction of Stock Market Trends with Twitter; Face Recognition in the Wild; Extracting text from Images; Mobile Logo Recognition; Developing a Web Service Framework for OpenIMAJ; Contextual and Lyrical Music Information Extraction; Generating 3D models using a Kinect Camera; Image Classification to Support a SenseCam System; Scene/Voice Detection in News Footage Using Audio/Visual Cues.
I developed and led our course in data-mining that that first ran in the second semester of the 2015/16 academic year. This module covers tools and techniques for modelling and mining data, with a particular focuses on unstructured data.
Together with Adam Prügel-Bennet, I'm part of the lecturing team for the advanced machine learning module. My lectures cover large-scale machine learning. I also help supervise the group projects undertaken by the students.
I was the module lead for the second-year web science group project module. In this module, I got the students to carry out a piece of research using a large dataset crawled from the web. One year I gave the students access to set of 46 million geolocated images and associated metadata from flickr and asked them to come up with their own research topics. One of the teams looked at whether a countries suicide rate could be predicted from the proportion for group photos to photos of individuals taken within that country and another looked at exploring whether photo tag sentiment correlated with indices of world happiness.
I delivered lectures on programming with Javascript and cloud computing in this module. I'm was also responsible for the individual coursework element, in which I get the students to develop a multi-user dungeon using Node.js. During the weeks when the coursework is running I ran regular drop-in sessions in the undergraduate computing laboratories for students to get help.
Supervised two GDPs: one on automatic video summarisation and the other on automatic image annotation.
Co-supervised a GDP project called "iSee: Intelligent Scene Extraction Engine" which looked at automated video analysis and indexing.
Delivered lectures on multimedia analysis and retrieval.
Delivered lectures on information retrieval and multimedia information handling.