Individual Project Ideas for Undergraduate and MScCS projects

 Staff: Email Graham Roberts if these details are incomplete or incorrect. 

 Students: Fill in the registration form when you have agreed a project with one of these supervisors. 

Projects with External Supervisors

These are interesting projects proposed by people external to the CS department. To do one of these projects you will also need an internal supervisor (one of the CS staff listed on the rest of this page). Contact the external supervisor using the contact details given in the details.

• BisoMotion. Using real time information to provide quicker and richer updates on the likelihood of events happening. More information here. Contact Joshua Wallace (joshua.wallace@bisomotion.co.uk) for further information.
• FlashSocial Ltd. FlashSOC is a visual platform that enables people to instantly capture and share the present, building a congruent stream of user perspective. There are two project ideas, one for web-based development and the other for iOS. More information here.
• Projects with Mpyxis Ltd. (mobile and secure transactions). See here for more information.
• A Modified Simplex Algorithm for Optimisation of Discreet Variables (MEng) with Spyridon Konstantinidis, Dept. of Biochemical Engineering. More information here. Contact Spyridon Konstantinidis s.konstantinidis@ucl.ac.uk
• Tipgain Mobile App Student Project. More information here.
• iProov Limited. Contact Andrew Bud andrew.bud@iproov.com if interested.
  • “iProov Limited, a London-based tech startup, is offering the following project, working closely with the CEO and Lead Developer.  iProov is developing a revolutionary way to verify the identity of an individual.  The system includes an app for a mobile device and some sophisticated processing in the cloud.  The device app has to manipulate a number of device APIs in an unusual way, and do so at high speed.  So far, all development has taken place on the Android platform.  The Undergraduate Project will invite the student to develop an Apple IOS5/6 app for the iPhone and iPad with the same functionality.  The challenge for the student will be to familiarise themself with the programming language, development environment and APIs of IOS, and to devise ways to implement the specified functionality within this context.  The Android environment is so different to IOS that the engineering involved is not a mere porting, but an interesting and demanding development from scratch.  The student will work in frequent contact with the Lead Developer, himself a top graduate of UCL CS, and with the CEO, a successful entrepreneur with 30 years of engineering and business experience.  There may be opportunities for subsequent employment.”
     
  • “iProov Limited, a London-based tech startup, is offering the following project, working closely with the CEO and Lead Developer. iProov is developing a revolutionary way to verify the identity of an individual.  One of the avenues iProov is exploring involves processing a stored video stream to extract and analyse certain characteristics.  These characteristics can be present with low signal-to-noise ratios and hence must be extracted using some creative image processing. Once extracted, these characteristics need to be recognised and characterised, a pattern recognition problem which may involve some machine learning.  The student will be expected to do a literature review to find any relevant academic research, and then to build a prototype with which performance tests will then be undertaken. The current environment includes OpenCV accessible through Python which should provide the necessary framework to develop the prototype. It is expected that a relevant portion of the project time will be spent on testing and trying out different approaches, to find one that works well.  The student will work in frequent contact with the Lead Developer, himself a top graduate of UCL CS, and with the CEO, a successful entrepreneur with 30 years of engineering and business experience. There may be opportunities for subsequent employment.”
• Footzie Ltd is a new London-based startup, which has created a unique football player trading game (www.footzie.co.uk). Users of the game build a portfolio of Premier League players within a budget of £100m, with the aim to increase the value of their portfolio each week. The prices of players are dependent on match ratings and market forces of supply & demand (i.e. when users buy and sell the player).
  
  Currently an initial webisite has been set up but the Undergraduate Project would involve automating and fine tuning back end processes. For example, we would like player ratings to be automatically extraced from SkySports. There are a few minor bugs with the market system, such as incorrect account values being displayed on occasion. We would also like to use the data from the back end to offer additional features, such as a ticker of prices of most bought players for instance and an addition of a monthly leaderboard.
  
  Taking on this project will enable you to get involved with a fast-growing and innovative start up. You will work closely with the Directors, as well as our designer and technical officer. You will need to develop a solid grasp of the programming language to implement some of the back end changes, and a creative mind would also help when trying to improve the design of current features.
  
  We look forward to hearing from you! (If you are interested and would like to find out more, please email info@footzie.co.uk)
• From Nadine Chadier (nadine.chadier@btinternet.com)
  
  • EMPOWERING FRENCH PRONUNCIATION: I have taken the sounds of the French
    language apart and coded it to empower language learners to pronounce French
    themselves. In order to learn the codes, I have animators bringing each
    sound for every language alive. The animations will be part of a learning
    package available to download from a webpage and become an app.
  • A FRANCOPHILE VIRTUAL COMMUNITY: I would like to create a web and app
    based language French language "maintenance" community that I would host for
    anyone who has some French language skills and wishes to practise, keep them
    up to date and maintain if not improve their standard without making the
    time or bearing the expense to take regular lessons. The members would
    report to me on the web or through an app on a daily/weekly or monthly
    basis, according to the commitment level they wish to have.

 Areas of Interest: image processing, machine vision, medical imaging and audio processing 

 Project Suggestions: 

• Visualization in diffusion MRI 
• Enhancement of Palimpsest Images 
• Image-based priors for diffusion MRI 
• Segmentation of the spinal cord 
• Parallel stochastic optimization 
• Weapons of moss destruction
   

See here for a list of project ideas.

Jade Alglave

Areas of Interest: Programming Principles, Logic and Verification

Please contact directly for project ideas.

 Areas of Interest: image processing, graphics, medical imaging 

 Project suggestions: 

• 3D Optical Tomography for Breast Cancer Detection 
• Meshing the human head from boundary surface measurements 
• Fast functional imaging of the brain using infrared light 
• Time Series Analysis Methods in Optical Topography 
• Solving the Radiative Transfer Equation with the Light Field Method 
• Multi-Spectral Probabilistic Diffusion 
• Stochastic Optimisation for boundary detection 
• Multimodality Image Reconstruction with Mutual Information Constraint 
• Multigrid methods for Finite Elements 
• Parametric surface methods for Finite Elements 
• Bubble Meshing 
• Parallel FEM on the Beowulf 
• Parametric shadows 

More details here. 

Tomaso Aste

Areas of Interest: Financial Computing

Please contact directly for project ideas.

David Barber 

 Interests: Machine learning, Bayesian methods. 

 Project Suggestions: 

•  See here for a list of suggestions.
  

  ---

Earl Barr

Areas of Interest: Software Engineering, Testing

Please contact directly for project ideas.

 Area of Interest: evolutionary computing 

 Project Sugggestions: 

• Fractal Proteins 
• Evolutionary Sculptures 
• Ant Colonies for Learning 
• Evolving Lindenmeyer Systems 
• Evolving Aerodynamic Shapes 

Nadia Berthouze 

Area of interest: affective computing, emotion recognition, user modeling 

 Project suggestions: 

• Recognition of human emotion in computer games 
• Recognition of human emotion in patients 
• Feedback mechanisms in a self-organizing neural network 

 Emotion recognition will be based on body gesture or biophysical
 changes in players/patients. 

 More details here. 

Gabriel Brostow 

 
 Areas of Interest: computer vision, animation, image processing, graphics 

 Project suggestions: 

• Food and exercise diaries for sports and fitness (different levels: therapy through Olympics)
• Video Metrology: measuring size and speed in the real world
• Building databases of photos or videos using games and clever user interaction
• Computational Photography for art and science, for example, time-lapse

See other suggestions here www.cs.ucl.ac.uk/staff/G.Brostow

Duncan Brumby 

 
 Areas of interest: Human-Computer Interaction, Information Search, Multitasking, Driver Distraction 

 Project Suggestions 

• 1. Dual-task interactions: How do people interleave attention when multitasking? One dominant account is that the completion of a subtask serves as a cue to switch tasks. But what happens if switching at subtask boundaries led to poor task performance? This project will focus on how people allocate resources between multiple ongoing tasks while driving. Experiments will be conducted in a desktop driving simulator using specially instrumented devices for secondary task interactions. The experiments will be informed by various computational accounts of how people might choose to schedule resources between tasks.
• 2. Eyes-free interactions: Project will focus on evaluating bespoke systems designed to minimise dual-task interference by making use of auditory output to reduce demand on the visual channel. One particular eyes-free system is Zhao et al.’s earPod, which makes use of synchronous audio feedback to support menu selection on mobile devices. But if we consider a typical portable digital music player, there might be several thousand songs represented in a hierarchical menu structure. Learning and navigating this structure could be cognitively demanding. It is not clear whether moving to eyes-free interaction will ameliorate the harmful effects of distraction because of the significant increase in cognitive load that these interactions require.
• 3. When to click? How do people decide when to stop searching a new web page and focus on selecting a link in pursuit of a particular goal? Is the position of an item in the page important? Do people look at all of the items on page? These are just some the questions that we have recently tried to investigate. I would be interested in supervising a project that develops a richer (computational) understanding of how people make moment-to-moment choices when engaged in an interactive search task. 

More details here

Kevin Bryson 

 Areas of Interest: Bioinformatics and Systems Biology (computational analysis of DNA sequences/structures, genomes, protein sequences/structures, microarrays, biological networks, amyloid diseases and bacterial biofilms).
 

 Other non-biological project areas supervised: Development of software for IT teaching and careers.
 

 Project suggestions here 

Licia Capra 

Area of Interest: Financial Computing 

Project Suggestions:

• If you are interested in understanding how the financial markets go wrong, you might be interested in "financial stability" as a topic for your final year individual project (3rdyr BSc and 4thyr MEng).
  
  The financial markets have been in turmoil for several years, and especially since the rise of High Frequency Trading (HFT).  That doesn't mean HFT is the cause of the problems - just that things are hotting up!  So how can we better understand what is going on?  My team is viewing the financial markets as huge distributed computer systems and investigating modes of instability - how to model them and understand them better.  We have started with the "Flash Crash" of May 6th 2010 (where the US markets went completely out of control for 30 minutes), and have explored in detail one particular part of the Flash Crash called the "Hot Potato" effect where about 15 HF traders traded furiously with each other in a bizarre panic which only stopped when a major exchange stopped all trading for 5 seconds.  Given that HF traders can issue tens of thousands of orders a second, that's a very long time!
  
  We have now managed to recreate the Hot Potato effect in the laboratory - we have developed an algorithmic model of how a number of trading algorithms can interact with each other and enter a destructive panic mode of co-oscillation; this model has been subjected to phase-space analysis, and we have shown how the system may be stable, may oscillate forever, or may engage in damped oscillation that dies out after a while.  An agent-based simulator has been written (using the functional language Miranda) and the simulation results show the same effects as the model; and we have also looked at the possibility of using the Pi Calculus to model feedback effects in the financial markets.
  
  I have already agreed to supervise one project in this area and have room for another if you rate yourself as an excellent programmer (with excellent English).    The precise project description will depend on your skills, but may for example be one of the following:
  
  1) further defining the new modeling language based on the pi calculus (some work has already been done on this);
  
  2) writing an abstract machine to make the modeling language executable;
  
  3) designing and implementing a graphical/visual front-end to the modeling language;
  
  4) using what has already been developed to investigate some other feedback loops in the markets (we have identified a large number of these) - to model them and to simulate them.
  
  
  If you are interested, please email clack@cs.ucl.ac.uk

David Clark

Areas of Interest: Software Engineering, Testing, Program Flow Security

Please contact directly for project ideas.

Nicolas Courtois 

 
 Area of interest: Information Security, Cryptography 

 Dr. Courtois did not provide specific project suggestions. Please contact him by e-mail if you want to do a project in this field. 

Ingemar Cox 

 
 Area of Interest: Media Futures 

 Project Suggestions: 

• 1. A comparison of document fingerprint algorithms. Document fingerprints are widely used to compare a document with a database of other documents, for the purpose of detecting plagiarism. A variety of different fingerprints have been proposed. Different algorithms have different free parameters that affect performance. This project would investigate the affect of these parameters on performance and compare the performance of various algorithms.
• 2. Data mining document citation information. Document fingerprints can be used as keys to access a database that provides complete citation information for the document. This database can be constructed in a variety of ways. This project will consider a data mining approach using three primary sources. The first is 2Tbytes of data related to CiteSeerX.com. The second is a combination of citation data from dblp together with pdf's retrieved via GoogleScholar.
• 3. Using shingles for detecting near-duplicate images. Shingles are a form of document fingerprint that are used to detect near-duplicate web pages. This project would investigate how to adapt shingles for the purpose of detecting near-duplicate images.

 Areas of Interest: human computer interaction, user interface design, human factors of information systems 

Ivana Drobnjak

Areas of Interest: computational modelling, biomedical imaging, diffusion modelling

Project Suggestions:

• Signal computation in diffusion MRI
• Using computational modelling to learn about tissue microstructure

 Area of Interest: software engineering, testing, computer music 

Project Suggestions:

• See here for a list of project ideas.

Denise Gorse 

 Area of Interest: bioinformatics, neural networks 

Dr. Gorse did not provide specific project suggestions. Please contact her by e-mail if you want to do a project in this field. 

Area of Interest: Human Vision, Computer Vision, Image Processing, Medical Imaging

Project Suggestions: 

• See here for a list of suggestions

 Area of Interest: Networking, especially security and mobile systems, ubiquitous computing. 

Specific areas of current interest: 

• The application of sensor systems to monitoring elite athletes. 
• Building wireless muscle oxygenation sensors 
• Signal processing of accelerometry information 
• Foot pressure sensors 
• User interfaces for coaches
   

 Area of Interest: Networking, especially routing. 

Project suggestions:

• IS-IS routing protocol implementation for XORP. 
• XORP on XEN 
• Web-based user interface for XORP. 
• Practically anything else in XORP you're interested in. 
• DCCP protocol implementation in FreeBSD or Linux. 
• Simulation study of XCP-lite. 
• IP over Two Tin Cans and a Piece of Wet String 

David Hawkes
 
 Area of Interest: Medical Imaging
 

 Professor Hawkes did not provide specific project suggestions. Please contact him by e-mail if you want to do a project in his field.
 

 Area of interest: theoretical machine learning, online learning algorithms, kernel methods, bioinformatics. 

 Project Suggestions: 

• Graph based learning 
• Web spam detection 
• Spectral Clustering (New methods, "Parallel Distributed methods") 
• Online learning (Exploration versus exploitation see pascal challenge) 
• Game playing (Hex, Poker, Go) 
• Collaborative Filtering
• Experimental or theory-based(difficult) projects in machine
   learning possible 

Derek Hill
 
 Area of Interest: Medical Imaging 

 Professor Hill did not provide specific project suggestions. Please contact him by e-mail if you want to do a project in his field.
 
 

 Area of interest: Temporal and Modal Logic, Algebraic Logic, Game Theory, Planning, Complexity 

 Area of interest: knowledge representation and reasoning; Argumentation 

 Project suggestions: 

• A system for generating first-order semantic tableaux.
• A system for automatically organizing sets of scientific articles in pdf. The system should be able to catalogue the papers, generate tags for indexing, relate articles by index, and by bibliographic reference.
• A system for sense making with a collection of points of view in a discussion. The system should be able to store, present, and analyse points of view for diverse discussions (e.g. should the UK Government support a new airport in the Thames estuary? should HS2 be built? should students have to pay a fee to study at university? should the NHS be privatised? Etc.). The system could use graphical and/or textual representations of arguments and counterarguments.
• A natural language understanding system for extracting and understanding textual information from websites. The system will be based on a toolkit such as the Python NLTK system for natural language processing, or the Java GATE system for text engineering, and it will be developed for a focussed domain that is of interest to the student (e.g. news, business, science, ecommerce, social networking, ...). The aim depends on the student’s interests, but it could be to do sentiment analysis (e.g. product or service reviews), or to populate a database with information gleaned from the web.
  
  

Kyle Jamieson 

 
 Area of Interest: Networking, especially wireless networks and digital communications 

Project suggestions:

• How to make your wireless connection more robust when you are in areas of spotty coverage.
• A study on radio spectrum availability in "TV whitespaces," areas of the radio spectrum freed up for data use because of the analogue to digital TV switchover currently in progress.
• Smartphone keyboard TEMPEST: snoop on the typing sounds of a nearby smartphone, figure out what the person typed.
• Localisation of GPS satellites in the sky using phased-array antennas, for the purpose of anti-spoofing and anti-jamming of the GPS signal.

Further information here.

 Area of interest: bioinformatics, protein structure prediction and analysis, simulations of protein folding, Hidden Markov Models, transmembrane protein analysis, machine learning applications in bioinformatics, de novo protein design methodology, and genome analysis including the application of intelligent software agents. 

Simon Julier 

 Area of interest: Tracking, computer vision, augmented reality, sensor fusion 

 Project suggestions: 

•  A detailed list of project ideas is available here

Brad Karp 

 Area of interest: Networks and computer systems, spanning wireless and sensor networks, Internet worm quarantine, and Internet-scale distributed systems. 

 More details here.
 

Jan Kautz 

 Areas of interest: realistic graphics, video- and image-based rendering 

 Project suggestions: 

•  A list of project suggestions is available here.

Jens Krinke

Areas of interest: Software Test, Program Analysis, Clone Detection, Bug Detection

Project suggestions:

• Search for how code is copied: Where is my code coming from and where does it end up?
• Multi-version code hashing and search.
• Find irregular patterns in source code and natural text for plagiarism detection.
• Identify reused code between Eclipse plugins.
• Detection of Changes in Public Interfaces
• A diff/patch combination that is aware of copy, move, and rename operations.

See here for more detailed information.

Email if interested in a project in one of these areas.

 Areas of interest: Requirements Engineering, Logic Applied to Software
 Engineering 

 Project Suggestions: 

• A Wiki-Based Tool for Goal-Oriented Requirements Engineering 
• A Temporal Logic Calculator for Users of Model-Checking Tools 
• Automated Verification of UML State Machine Diagrams 
• A Framework for Developing Virtual Environments from Requirements
   Models (joint supervision with Simon Julier)
   

Zhaoping Li 

 Area of interest: models and behavorial experiments on human vision. 

 Project suggestions: 

• Eye tracking of human subjects in visual search tasks 
• Eye tracking of human subjects in face detection task 
• Laptop implementation of visual psychophysics experiments for experimenting in the public domain. 
• Measuring the speed of image processing in humans. 
• Projects of various sorts (to be discussed) on modeling human vision or visual inputs. 
• Various other behavioral experiments on human vision to be discussed. 

Niloy Mitra

Areas of interest: interactive images, high-level structure extraction from 3D points sets (e.g., from
Kinect, etc. scanners), 3D scene modeling, fabrication-aware Lego designer

Related projects can be seen at www.cs.ucl.ac.uk/staff/n.mitra/research/. For further details,
please directly contact Niloy (n.mitra@cs.ucl.ac.uk).

Dean Mohamedally

Area of interest: Constructionism in software engineering, knowledge elicitation and expertise transfer, NUI interaction, cross platform architectures, "Apps" development frameworks and templates.

Project Suggestions:

• MICROSOFT UK-SUPPORTED PROJECTS - Windows 8 Tablet and Table (Surface 1.0 and PixelSense) based apps for software engineering methods, software project management and business process modelling.
• Robotics and Drones projects using .Net Gadgeteer platform and GPRS/GSM/3G communications.
• Cross platform build toolkits for improving porting strategies between Windows 8 .Net and Android/Java. For example, user interfaces layouts, database and file processing methods, graphics calls. (C#, .NET MSIL  + Java/Android) - co-supervised with Dr Earl Barr.
• Kinect based tools for applying interactive constructionism to SE methods (XNA/C#/C++)./C#/C++)
• HTML5/AJAX widget toolkits for rapid web application development supporting NUIs and touchscreens (HTML5/AJAX/PHP)
• Windows 8 and Android hybrid app projects that specify a library API construction specific to a non-CS domain (e.g. education/training, imaging/photography, medicine/architecture).
• Office and in-home automation technologies (either linux custom builds, C/C++/Arduino, or Windows Embedded with .Net Gadgeteer platform).

Wole Oyekoya 

 Area of interest: Learning in 3D, Virtual Reality

Project suggestions: 

• See here for a list of ideas.

Massi Pontil 

 Area of interest: Interested in machine learning theory, pattern recognition and statistics, machine learning problems arising in computational vision, natural language processing and bioinformatics. 

• Multiple-output regression and spaces of vector-valued functions 
• Object detection in images with spectral methods 
• Making kernels for graphs 
• Learning object parts and hierarchical classification 
• Stability of bagging and boosting 
• Learning from multiple sources 
• Online algorithms for learning preference models 
• Comparison between different feature selection techniques 

 Area of interest: Dynamic programming languages, Java, web-applications, testing and test-driven development.

Graham is interested in projects involving: 

• iOS and Android Application Mobile Development
• Use of devices such as Raspberry Pi, Arduino, .NET Gadgeteer, Kinect to build robots, sensing applications, wearable devices.
  • Build a 3D scanner using Kinects.
• Application development using dynamic languages like Groovy, Ruby
   and Python. 
• Web application development using frameworks like Rails, Grails orJSF/Ajax. 
• Testing techniques and the development of testing tools. 
• Java application development. 
• A Moodle 2 offline quiz editor, that can run and sync on multiple platforms.
• A question/response/clicker system that can be used in lectures, where students use mobile devices to
  respond in real time to questions, and the results are collected and displayed.
• An entire workflow for the electronic marking of coursework submitted via Moodle.
• A home automation and security system that can be controlled remotely, built using open source software and parts.

Other project suggestions are here. 

Angela Sasse and Martin Ruskov

Development of User Interface for a Serious Game about Information Security

Supervisors: Martin Ruskov, Prof. M. Angela Sasse

This project will consist of the development and evaluation of the user interface for a prototype of a serious game. The interface needs to allow for asynchronous interaction between several users. Core requirement for the interface is to provide on-demand information and incremental increase in complexity visible to users.

The game has the objective to improve awareness and deepen understanding of users about one particular theory of user compliance. The theory being explored is about people's limited ability to comply to requirements and constraints that are not in the primary focus of their work.

Tasks

Literature review on online games and multiplayer games

• Literature review on interaction design and usability evaluation
• Development of prototype system with web technologies
• Design of software evaluation
• Carry out experiments and analyse results

Requirements

• Interest in interaction design
• Experience with web development (both client- and server-side) is a plus

Learning Outcomes

• Understanding of theories of usability and learning
• Experience with computer-mediated interfaces
• Experience with software evaluation
• Experience with agile software development

Contact M.Ruskov@cs.ucl.ac.uk for initial enquiries.

Development of Game Model for a Serious Game about Information Security

Supervisors: Martin Ruskov, Prof. M. Angela Sasse

This project will consist of the development and evaluation of the object model and business logic for a serious game. The model needs to allow for the analysis of qualitative data provided by users. Essential to the project is that the system can provide relevant feedback to any given user input. Unlike other data analytic tools, here focus is not on accuracy, but on enabling a supervised machine learning paradigm that provides value to users training the system.

The game has the objective to improve awareness and deepen understanding of users about one particular theory of user compliance. The theory being explored is about people's limited ability to comply to requirements and constraints that are not in the primary focus of their work.

Tasks

• Literature review on online games and multiplayer games
• Literature review on knowledge management
• Literature review on data and text mining
• Development of prototype system with web technologies
• Design of software evaluation
• Carry out experiments and analyse results

Requirements

• Interest in data mining
• Experience with web development (both client- and server-side) is a plus

Learning Outcomes

• Understanding of theories of knowledge management and data mining
• Experience with information systems
• Experience with software evaluation
• Experience with agile software development

Contact M.Ruskov@cs.ucl.ac.uk for initial enquiries.

Human-centred technology/service design

 Project Suggestions: 

• Targeted-advertising visualising web-browser plugin
  
  Miguel Malheiros
  
  Users are increasingly being targeted by advertisement tailored to their personal tastes. In order to this the advertising organisations, such as Google, record users’ online activities – including the searches they make, the pages they visit, the content they view, and their interactions on social networking sites – in a ever expanding profile that can be processed to create and display more personal ads.
  
  Previous studies we have done in our group show that most users do not understand how targeted advertising works, and as a results, they don’t have privacy concerns. Whens shown what information is collected and how it is used to target ads, many react quite differently.
  
  The main goal of the project is to develop a web-browser plugin that enables the visualisation of personal information flow around targeted advertisement. This tool should be able to detect when the user is being shown targeted ads as well as detect when the user’s personal and behavioural data is being collected for later use in the creation of targeted ads.
  
  We have a detailed description of how Google tracks users.  The plugin can be developed for any browser and in any programming language the student is comfortable with, but our the preference would be for a Firefox plugin. If there is enough time, the second goal is to carry out a usabilty evaluation of the plugin (the deign of the study would be developed in collaboration with usability researchers.) The plugin must be easy to install by non-expert users on their home computer. (This will help to conduct the evaluation with a large number of users.)
  
  The project would suit a student with good programming skills who is looking for a challenging project that is part of an ongoing research programme.  An interest in usability and privacy issues helps.
  
  

• Improving Road Safety by Detecting Driver Distraction

Supervisors: Ashweeni Beeharee/Angela Sasse

This project will investigate how driver distraction can be detected and
what measures can be taken to draw the drivers attention back to driving.
Driver distraction is one of the most significant causes of road
accidents. The detection of driver alertness will use information that can
be readily obtained from the on-board computers in vehicles  which is a
very innovative method. This study will interface with ongoing work on
driver distraction within the SafeTRIP project.

Tasks:
    Literature review on driver distraction and detection methods,
OBD, CAN
    Development of application to detect driver distraction
    Development of application to remedy driver distraction (e.g.
alerting driver, alerting authorities, etc)
    Design and carry out experiments to evaluate effectiveness of the
application

Skills:
Programming skills in Java or C#

Learning Outcomes:
    Understanding of driver distraction, road safety
    Experience in developing real-world road safety application
    Experience working on a project with real impact on the motoring
industry

Contact A.Beeharee@cs.ucl.ac.uk for initial enquiries.

• Visual Attention on Tablets

Supervisors: Ashweeni Beeharee

This project will explore investigate if change blindness occurs on tablet
PCs (such as Motorola Xoom and the iPad). Change blindness is the
inability to see change in a scene that we are looking at, though it may
be something significant. From visualisation and information presentation
point of view this is an interesting question and it is the subject of
much research work in games industry. As tablets can run fairly complex
applications ranging from simple text application to augmented reality
games, it is possible that certain information presented to the user is
missed out or go unseen.

Tasks:
    Literature review on change blindness in visualisation
    Development of the application on a suitable platform (student to
evaluate suitability). There is also the possibility to modify  an
existing tablet application
    Deploying it on actual phones
    Carry out experiments with users.

Skills:
Programming skills in Java or C#/C++

Learning Outcome:
    Understanding the impact of visual perception and change blindness
on visualisation
    Gain experience writing and deploying mobile application
    Acquire knowledge of development on Android platform
(phone/tablet) or optionally iOS

Contact A.Beeharee@cs.ucl.ac.uk for initial enquiries.

• Development and Evaluation of Pragmatically Enriched Chat
  
  Supervisors: Martin Ruskov
  
  This project will consist of the development and evaluation of a web-based idea discussion system, expanded with pragmatic annotations.
  The development of an enriched web chat with interactions modelled after the simulation game vLeader www.youtube.com/watch. The two features beyond traditional chat systems are ideas and pragmatic annotations. Upon submission each message (or utterance) needs to be annotated by its author with a simplified quantitative model of intention in a conversation (see below). The other required feature is an object representation of a certain set of ideas. Ideas are potential objects of discussion, which could be explicitly addressed as part of the annotation. Whereas utterances are of key importance for a discussion between people, the annotated intention is the one being interpreted by automated agents.
  
  The annotation model of pragmatics of a message (as adopted from vLeader) is required to feature:
  •    target – who or what is this being meant as object of intention of this utterance. This could be either a person or an idea
  •    valence – strength of intention on the interval [-1;1], the sign representing positive or negative  intention towards  the target
  
  Tasks
  •    Literature review on computer-supported collaborative work
  •    Literature review on usability evaluation
  •    Development of prototype system
  •    Design of software evaluation
  •    Carry out experiments and analyse results
  
  Requirements
  •    Experience with web development (both client- and server-side) is a plus
  
  Learning Outcomes
  •    Understanding of theories of conversational pragmatics
  •    Experience with computer-mediated communication systems
  •    Experience with software evaluation
  Contact M.Ruskov@cs.ucl.ac.uk for initial enquiries.

 John Shawe Taylor 

 Area of Interest: Machine Learning: Theory and Applications 

• Collaborative filtering using the margin perceptron algorithm 
• Targeted probability density learning 
• Feature selection methods exploiting time coherence of images 
• Canonical Correlation Analysis for images and text 
• Transduction using graph theoretic algorithms 
• Analysis of brain scans of subjects listening to music 
• Machine learning techniques for object detection 
• Semantically informed document analysis 

 More details here. 

Dan Stoyanov

Areas of interest: Medical Image Computing, Computer Vision, Computational Stereo, Scene Flow, Structure-from-Motion, Robotic Surgery, Surgical Skill Analysis, Biophotonics

A page with project suggestions is here: www.cs.ucl.ac.uk/staff/Dan.Stoyanov/teaching.html

 Project Suggestions: 

 I am interested in all areas of immersive and non-immersive virtual reality systems. This includes real-time rendering, model generation,
 systems programming, distributed application building, collaborative interaction and usability engineering for VE systems and presence. 

 I am particularly interested at the moment in students wanting to do experiments with our virtual reality systems: 

• Reproducing the "rubber hand illusion" inside a virtual reality system. 
• Evaluating the impact of avatar eye gaze and lip synchronisation in simulations of stressful situations. 
• Running our "fear of public speaking in virtual reality" experiments, in different types of mixed-reality systems. 
• The impact of live and simulated animation on perception of avatars. 
• Understanding distortions of perceptual space when embodied in an immersive virtual environments. 

For specific project details, see here.

www.cs.ucl.ac.uk/staff/A.Steed/teaching/projects.html 

 Area of interest: intelligent systems, data mining. 

• Web Site for UltimateModels 
• Target Rifle Simulator 

Jun Wang 

 Area of interest: Information Retrieval and Data Mining 

 Project Suggestions: 

• Mobile location-aware search. Consider now you are in London and want to find a restaurant nearby. Can your smartphone suggest one, based on your location and navigate you through the road? Can you also access and get a suggestion from other people who have already visited to the restaurant? Can Twitter help us find a solution from word-of-mouth? 
• Web Sentence Search. Have you experienced any difficulties while you are writing your English essays? If yes, the Internet might help you. Tons of web pages from many online news paper websites provide most authoritative examples for English writings. But the problem is how to find them. More specifically, how to retrieve examples of sentences or paragraphs that contain the similar usage of the term that you are looking for. Be warned about plagiarism though. 
• Economic models for online information access. Financial models are advanced in modelling risk. In Information Retrieval research, we have identified the analogy between Web search markets and Stock Markets, having developed novel methods to handle retrieval risk and search result diversification. For more information, check out our recent studies. But this is just a start… 
• Financial data mining Online materials: Data mining, Teaching Financial Data Mining, Data Mining for Financial Applications. 
• Items (movies, music, books, news, web pages, etc.) that are likely of interest to a given user. As one of the common techniques, collaborative filtering makes recommendation on the basis of like-minded people. Popular systems include those offered by Amazon, Netflix, LastFM, etc. Although collaborative filtering is an effective way to alleviate information overload, it may be vulnerable to recommendation attacks since it solely relies on the item ratings from users. Attackers, who intend to have their products (items) recommended more often, introduce biased ratings (preferences) to influence recommendation predictions. This project will explore various attack methods, and particularly study the effectiveness of these attacks towards different collaborative filtering algorithms. As a result, we expect to have a new way to evaluate and detect these attacks on recommender systems. 

 More information can be found at: web4.cs.ucl.ac.uk/staff/jun.wang/blog/teaching/projects/

Shi Zhou

Area of Interest: Internet, Complex Networks, Network Security, Social Networks, Transport Networks

Project Suggestions:

• See the list here.