Augmenting vision by combining augmented reality with models of visual attention

Augmenting vision by combining augmented reality with models of visual attention.

The human visual system is remarkable. It can single out objects quickly and monitor our surroundings. Yet, seeing changes between two mostly identical scenes is difficult and critical visual information is sometimes missed. We propose to use augmented reality (AR) to remedy these limitations. In AR, people wear a headset that lets them see the real world while a computer system overlays information or substitutes part of the real world. However, current AR systems are not based on psychological knowledge about vision. Thus, the PIs of the project join expertise on AR and visual attention. The project will validate how well existing models of visual attention can be used in AR. We also develop new experimental paradigms for measuring and modeling visual attention to extend the short experiments on visual attention to realistic tasks. We use the resulting insights on attention to prototype AR systems that improve vision, including finding objects and detecting changes in complex scenes. Finally, we evaluate the usability of the prototypes for end users and their support for visual problem solving. The outcome of the project is demonstrators of augmented vision, as well as basic research in AR and models of attention.