LUMAUS - Research project

 

 

Concurrent with the increase in the average age of people, growth in the number of persons with low vision is unfortunately expected. It is estimated that the number of visually impaired people could triple in the next ten years. Low vision can seriously affect the ability to perform simple activities of everyday life such as cooking or walking in security even at home. The quality of life can be dramatically altered and it can be impossible to maintain independence in a safe manner. Yet loss of independence is a predominant concern of the older adult.
All recent studies highlight a clear lack of efficiency and practicality of visual aids available on the market or developed in laboratories. To respond effectively to the societal challenge of assisting in their everyday lives people facing the development of a visual deficiency, new research investigations that break with the existing ones must be explored. The aim of the LUMAUS project is to exploit the potential offered by light combined with digital tools, such as Artificial Intelligence (AI), Computer Science and Mixed Reality (MR), for designing efficient visual aids. The assistive technologies must adapt to the user needs and expectations as far as possible to optimize the observation conditions of a scene for improving object detection, tracking and classification. This is an essential criterion for preserving the autonomy of the greatest number of low vision people, which is a real socio-economic public health issue.
The research works that will be carried out aim to develop predictive models for controlling smart adaptive lighting systems to change the environmental cues with the purpose to selectively enhance the residual vision of visually impaired individuals. Such models will be defined using data derived from psychophysical experiments integrating motion tracking. Panels of participants will include people with different types of visual impairment as well as people in low vision situations through a mixed reality simulation. The ultimate question is to know how to dynamically adapt the light characteristics and its 3D spatial distribution to the visual context and to a given pathology.