Multimodal kinect-supported interaction for visually impaired users

This paper discusses Kreader, a proof-of-concept for a new interface for blind or visually impaired users to have text read to them. We use the Kinect device to track the users body. All feedback is presented with auditory cues, while a minimal visual interface can be turned on optionally. Interface elements are organized in a list manner and placed ego-centric, in relation to the user's body. Moving around in the room does not change the element's location. Hence visually impaired users can utilize their "body-sense" to find elements. Two test sessions were used to evaluate Kreader. We think the results are encouraging and provide a solid foundation for future research into such an interface, that can be navigated by sighted and visually impaired users.