The colours of our world tell us many things - if fruit is ready to eat, if meat is sufficiently cooked, if someone has a rash or sunburn - but approximately 10% of all people cannot rely on colour signals because they have congenital, acquired, or situationally-induced colour vision deficiency (CVD - commonly called colour blindness). CVD has a wide range of severities - from subtle variations in colour naming to only seeing shades of grey - and can cause problems in digital interfaces ranging from minor nuisances (e.g., being unable to distinguish 'visited' from 'not visited' website links) to major safety concerns (e.g., misreading GPS navigation devices).

Recently, recolouring tools have been developed that replace problem colours with more differentiable colours in order to make interfaces more accessible for people with CVD. However, these tools are limited to individuals with dichromatic CVD - a particularly severe and somewhat rare form of congenital CVD. As a result, the bulk of people with CVD (those with acquired and situationally-induced CVD as well as those with non-dichromatic forms of congenital CVD) continue to have difficulties because current recolouring tools perform too much or not enough recolouring. In addition, current recolouring tools require the type of CVD to accommodate - which most people with CVD do not know - often leading recolouring tools to perform the wrong type of recolouring.

In this seminar, I will present my PhD research that extends recolouring tools to any user with any type of CVD by developing a model of colour differentiation that is sensitive to every individual and environmental factor that influences colour perception.

I will first discuss how this model efficiently measures and represents individual colour differentiation abilities. I will then present results from a recent comparative evaluation between a recolouring tool built using this model and existing recolouring tools. Time pending, I will also discuss additional extensions and applications of this research.