Research at UCL Computer Science

The department is brought together by a clear focus on experimental computer science; that is we are concerned with real-world data and observations about computer and software systems and their use. We develop theory that leads to hypotheses that can be checked experimentally, and where experimentation delivers results forming a basis for theory. This shared experimental orientation, which emphasises measurement, methodological rigour and ultimately reproducibility, not only binds together our research but determines our strategy and policy towards recruitment, directs our investment and underpins our research education. UCL Computer Science is a global leader in research in experimental computer science.

Research Groups

Research Groups are strong, broad-based, disciplinary teams associated with laboratories. These groups are formally established and constitute the focus for strategic development within the department.

Research Students

UCL Computer Science supports a large, active postgraduate community studying on both research and taught programmes. Please follow the links below for information on these programmes:

Interest Groups

There are also several Interest Groups which are informal groupings that come together to explore new subjects, extend the range of research or bring together different strands of computer science. Such teams reflect our dynamic research culture and are the seeds from which new Research Groups can grow.

Impact Stories

Our latest research case studies demonstrate the wide impact that our work has across society - from computational finance in The City; to groundbreaking medical imaging in the NHS; to information security in commerce and government. Read our case studies and see our academics talking about their work below.

3D Body Scanning: fashion, health, sport

3D Body Scanning: fashion, health, sport

Prof Philip Treleaven

UCL Computer Science has pioneered the use of 3D body scanning for national sizing survey (e.g. SizeUK, SizeUSA, SizeTHAILAND), clothing, healthcare and sport, and has launched three start-ups. SizeUK involved measuring 11,000 UK adults working with fourteen major retailers.

UCL and its partners developed company/brand specific size charts for over 25 UK retailers based on the retailer’s/brand’s specific demographic customer profile. Collaborative commercial partners included Boots, Bupa Healthcare and Tottenham Hotspur Football Club.

Prof Treleaven and his collaborators also developed a 3D healthcare system for GP practices that analyses and comments upon a patient’s 3D anthropometrics, comparing the patient’s scan to the SizeUK data.

This can identify any deviation of the patient from national norms, such as over or underweight and the shape of the stomach which is an indicator of the propensity for the onset of Type II Diabetes.

HeliconHeart: A clinical management service for stroke prevention

HeliconHeart: A clinical management service for stroke prevention

Prof Dipak Kalra, Prof  David Patterson

HeliconHeart builds on two decades of research by UCL’s Centre for Health Informatics and Multiprofessional Education (CHIME). It has been designed to meet specific needs:

· Clinically useful Electronic Health Record, standards-based

· Decision support: which drugs to use, how and when?

· Web software

· Governance & data analytics

· Collaboration tools

· Education

Who benefits?

· Payers/Commissioners: better outcomes through new service delivery models,  education and quality monitoring

· Clinicians: evidence-based best practice & continuing professional development, easy & secure communication with colleagues

· Patients: Empowerment, direct involvement, self-testing and self-management; managing their health whilst on the move

ArrayTrack: Precise real-time indoor localization from existing WiFi infrastructure

ArrayTrack: Precise real-time indoor localization from existing WiFi infrastructure

Dr Kyle Jamieson

Why is localization useful? For businesses: proximity triggered product information;  precise  location trails for customer analytics. For users: enabling precise augmented reality indoors; local gadgets (find my phone) and people.

How WiFi localization works: using angle of arrival information, we calculate a client’s bearing from a single transmission. With several access points we can triangulate the client’s precise location in real-time.

Why is localization difficult? WiFi operates in environments with obstacles and multipath reflections. Transmissions bounce off walls and the line of sight maybe weaker than multipath signals. Different access points will also see the incoming signal vary over time as the multipath environment changes.

Making it work in practice: using parallel processing on the GPU, we generate a heatmap for the transmission. With additional signal processing and filtering, we can achieve estimates with an error of ~23cm.

Our prototype uses off-the-shelf wireless equipment, and software is built to support next generation WiFi such as 802.11ac.

Camino diffusion MRI toolkit: microstructure imaging and connectivity mapping to avoid cognitive deficits after neurosurgery

Camino diffusion MRI toolkit: microstructure imaging and connectivity mapping to avoid cognitive deficits after neurosurgery

Prof Danny Alexander

Diffusion MRI is a specialised technique using the Brownian motion of water molecules to probe tissue microstructure non-invasively and effectively.

The technique is particularly sensitive to the fibrous architecture of white matter – the brain’s communication network. In combination with image-processing algorithms called tractography, diffusion MRI lets us visualize the connectivity of a living brain.

The Camino toolkit makes these algorithms freely available to  medical researchers and practitioners. For example, neurosurgeons use it to avoid damage to essential communication pathways while performing interventional neurosurgery. This helps avoid leaving patients with cognitive problems (such as impaired vision or motor skills), which can result from collateral damage during surgery.

Human-centred security in government and commercial applications

Human-centred security in government and commercial applications

Prof Angela Sasse

Prof Sasse created, developed and delivered the user-centric perspective that now underpins security thinking in both corporate and public-sector domains.

Compliance Budget: friction between information security and business processes reduces both security compliance and personal and organisational productivity. The user's ability to comply – the “compliance budget” – is limited. UCL’s research suggests security policies should be designed using human-computer interaction (HCI) principles to make it easier for users to ‘do the right thing’.

Balance Security, Productivity, and Cost: quantitative and qualitative data collection and analysis methods were used to obtain evidence of the effectiveness of security measures in corporate envronments.

Explicit authentication creates a “Wall of Disruption”: in a project commissioned by US National Institute of Standards and Technology, evidence was collected of the productivity losses caused by outdated explicit authentication mechanisms - the ‘wall of disruption’.

HP has incorporated the compliance budget model into its Security Analytics product. Prof Sasse’s work also underpins new and improved security products, including First Cyber Security’s SOLID and Safe Shop Window tools, which protects over 70% of UK online shopping revenue.

Improving prostate cancer diagnosis and care using computer simulation and medical image registration

Improving prostate cancer diagnosis and care using computer simulation and medical image registration

Dr Dean Barrett

 · This research has led to the introduction of technology to enable less invasive, image-directed strategies for treating and diagnosing prostate cancer, which is the most common cancer in men in the UK.

· It has led to the development of clinical criteria for new types of needle biopsy for the detection of prostate cancer.

· The computing technologies employed in the research include:

Biomechanical and statistical modelling of organ shape;

A novel image registration method for aligning geometric models with images;

Computer simulation of surgical procedures, such as needle biopsy, using image-based anatomical data.

New biopsy criteria are now used routinely to classify patient risk at University College Hospital where, since 2009, clinicians have determined the treatment options for more than 741 prostate cancer patients. The scheme has been adopted by a number of other hospitals locally and internationally, where its become the recommended standard of care.

Left: Computer simulation of transrectal ultrasound guided prostate biopsy.

Enhanced photo and special effects processing for professional and amateur photographers

Enhanced photo and special effects processing for professional and amateur photographers

Prof Jan Kautz

Prof Kautz and his team have developed two photo manipulation and processing methods (Exposure Fusion and Local Laplacian filtering) that are used to produce well-exposed photographs with tuneable local contrast.

Exposure Fusion is now considered the standard method for blending multiple   photographs into a single well-exposed photograph, and is used by a large number of commercial and non-commercial products.

Local Laplacian filtering was chosen by Adobe Systems Incorporated to be the default tool for image enhancements in Adobe Lightroom and Adobe Camera Raw.

As a result, these methods are now in the hands of hundreds of thousands users, who use them to create and manipulate well-exposed digital photographs.

xlinkit for fast, cheap, reliable banking with automated verification of over-the counter derivatives trading

xlinkit for fast, cheap, reliable banking with automated verification of over-the counter derivatives trading

Prof Wolfgang Emmerich, Prof Anthony Finkelstein

The Software Systems Engineering Group at UCL developed and patented xlinkit, an approach that supports the validation of XML documents in general and over-the-counter (OTC) derivative transactions expressed in the Financial Products Markup   Language (FpML) in particular.

Since 2008, the adoption of FpML by the financial services sector has become widespread, with 95% of the financial market participants who deal in over-the-counter derivatives now using it at some stage in their trading process, benefiting from reduced costs and risks in their trading.

The current Version 5.3 of the FpML Standard that is available for these market participants to use free of charge was released in October 2012. It now defines some 500 validation rules for a large number of equity, interest rate, credit, energy and foreign exchange derivatives.