Computer Science Project List 2017

Background

I am a senior lecturer in UCLIC. Therefore, my research projects involve understanding how people interact with technologies. I specialise in accessibility and assistive technologies and am particularly interested in designing for resource-poor settings.  A lot of my work is with disabled people and groups which help disabled people. Many of my projects will have come directly from people’s requests and you can expect to work with disabled people if you are doing a project with me.

A far as I am concerned designing for extreme user groups such as people who have no sight or are unable to hear  will make for better technologies for everyone. Much of my research involves experiments – you need to want to do stuff. Make sensor systems, attach them to people, measure, run algorithms, write code, make results accessible and do that  all over again.

I also think you should have fun and enjoy your project. Learning and discovering new things is after all exciting.

I have listed some example projects, but feel free to come up with your own and email me.

Students who are pro-active and who want to solve difficult, real-life problems should apply.

Project List

Media representations of disability

  • Create a data scraper for a variety of online content and then analyse this using Natural language Processing libraries such as spaCy to categorise the content and identify sources of media and/or event which cause a spike in more negative or positive coverage of disability. An example research question could be: Did the Olympics in 2012 create more positive word associations with disability in newspaper X compared to newspaper Y?

Developing World Projects

The  Evidence Collector

  • Globally, only 15% of people have access to assistive technologies such as eye glasses and wheelchairs. It’s difficult to provide evidence of just how life changing these technologies are for people on a global scale. Therefore there is a need for a  robust data collection tool for rehabilitation professionals especially in rural areas to collect and record patient achievements with their assistive technologies, while also helping to provide information to patients. Rural areas in places such as Tanzania have unreliable Internet and power, patients can have limited literacy. Therefore the interface needs to be simple but the system robust.

Virtual Wheelchair Skills TrainerDesign

  • Build and test a low-cost wheelchair training system. the system should be able to show people wheelchair skills and give feedback on their progress. This project will involve hardware and software design, 3D printing and biomechancal analysis and visualisation. The system will need to be developed to work in low-resource settings such as rural India.  Therefore, while the calculations may be complex the UX needs to be simple and intuitive to use.

NHS Projects

Detecting Dementia Decline

  • Dementia and detecting deterioration in conditions using a single body-worn sensor. We have a large data set of IMU date and people with different levels and types of dementia. This project would involve designing a visualiser for the data and algorithms to detect level of dementia. The project involves software and algorithm development including machine learning. This project is in conjunction with UCL’s Demential Research Institute.

Upper limb rehabilitation – measuring quantity and quality of movement

  • In collaboration with Nick Ward. This project aims to develop a system for measuring upper limb movement in a clinical environment and which includes feedback to both patients and clinicians which is easy to understand for them. This would develop the ARCCS IMU system which has been developed in-house by UCL CS. The goals of the project are to:
  1. Work with the clinical team to develop key metrics of success of the measurement system i.e. overall activity, quality of movement
  2. Develop algorithms to detect metrics automatically to distinguish level of movement and quality of movement.
  3. Test the system on a non-patient group to demonstrate the ability to detect the metrics.
  4. Collect single patient for 1 week in a clinical environment
  5. Investigate real-time feedback mechanisms for bot patients and students

 Open pose Projects

I am also interested in project proposals that make use of the Open Pose library for:

  • Rehabilitation projects
  • Crowd analytics.g. detecting disabled people or robots in a crowded environment
  • Accessibility projects e.g. sign language translator