Data Driven Tactile Displays for Enhanced Perception of Graphical Information

Tactile perception of graphical information has offered great potential in creating sensory substitution for the benefit of various applications. For example, blind and visually impaired (BVI) individuals have limited access to graphical information – photographs, diagrams, maps etc. Translating visual information into pictures to be touched via tactile images helps relieve this problem. However, existing assistive devices have received minimal user acceptance within BVI communities [1]. This is mostly due to the psychophysical deficiencies of expensive assistive devices and complicated training processes that severely limit the success rates in accurate tactile perception.

In this PhD project we will develop neural network-based algorithms to improve the performance of our patented high-resolution tactile display. Hence, we will design data-driven controllers to regulate the tactile sensations generated by actuators. This approach demonstrates high potential to tailor the psychophysics of these devices for individual users in different age and gender groups. Hence, we expect the AI-based controllers trained by the data collected from tactile behaviour of individual users to significantly improve the success rate for tactile perception. Therefore this project will enhance the graphical perception to promote accountable and responsible decision making among the BVI individuals using these systems.

In addition it will protect the BVI users by developing neural networks at different levels of abstraction and a more reliable platform for transparent data collection from individual users on personal assistive devices.

Researchers from the Departments of Psychology, Computer Science and Electrical Engineering collaborate at the University of Bath to develop and implement AI control algorithms for reliable tactile display devices. This PhD project will 

·        Develop intelligent control algorithms including neural networks for smart controllers that improve the system psychophysics for tactile behaviour of individual users. This high performance assistive technology will promote accountability of decisions made by users based on tactile perception.

·        Promote transparency on what type of data is collected from the users by developing multi-layer architectures that perform control and perception at different levels of abstraction tailored to individual user groups.

We have developed a new vibrotactile display device at the Faculty of Engineering [2] and [3], which has been tested for user perception at the Department of Psychology [4]. The supervisory team will provide comprehensive training and interdisciplinary skills and expertise in neural networks, machine learning, the psychology of perception, and electronic interface circuits.

The PhD student will have access to the labs of the Faculty of Engineering including the Microsystems Research Lab (https://people.bath.ac.uk/am3151/) and REality and Virtual Environments Augmentation Labs (http://www.bath.ac.uk/reveal). The PhD student will travel to national and international conferences to present the results of this interdisciplinary research to stake holders such as RNIB and end-users. The project will be carried out as part of an interdisciplinary integrated PhD in the UKRI Centre for Doctoral Training in Accountable, Responsible and Transparent AI (ART-AI). The ART-AI CDT aims at producing interdisciplinary graduates who can act as leaders and innovators with the knowledge to make the right decisions on what is possible, what is desirable, and how AI can be ethically, safely and effectively deployed. We value people from different life experiences with a passion for research. The CDT's mission is to graduate diverse specialists with perspectives who can go out in the world and make a difference.

Applicants are expected to have Masters or MEng degree in Computer Science, Electrical or Mechanical Engineering.

Formal applications should include a research proposal and be made via the University of Bath’s online application form. Enquiries about the application process should be sent to art-ai-applications@bath.ac.uk. Enquiries about the research should be directed to Dr Mohammadi.

Start date: 2 October 2023.