DESIGN AND DEVELOPMENT OF LARGE SCALE DATA COLLECTION FOR EYES FREE TEXT ENTRY

Abstract

Entering text on a touchscreen is challenging when users are unable to receive visual feedback due to their situation or disability. We are working on a recognition-based approach that attempts to infer the user’s intended text from a sequence of noisy tap data. We need to gather large amounts of such eyes-free data to develop and test our recognizer. We first interviewed users in the low vision and blind community to explore the best ways to engage users in our data collection effort. With the information gathered in the interviews, we are developing a cross-platform data collection interface that
will enable us to reach a large population of users who are visually-impaired.  We will describe our current progress in the development of our eyes-free data collector.