Touch input is generally considered inaccurate when it comes to small UI controls, such as keys on a keyboard or targets on a map application. While this has often been attributed to the so-called fat finger problem, I show that this is not the case and that touch is indeed accurate. I present the generalized perceived input model, which models touch input as a six dimensional operation and explains 67% of the touch inaccuracy. I also present two devices that exploit the new model in order to improve touch accuracy: a fingerprint scanner that increases the accuracy of current devices by nearly two times and an optical tracking system that improves accuracy by over three times. This increase in accuracy allows making touch interfaces more reliable, putting over 10 times more controls into the same user interface, or bringing touch input to very small mobile devices.
Bio: Christian Holz is a PhD student in Human-Computer Interaction working with Patrick Baudisch at Hasso Plattner Institute at the University of Potsdam in Germany. His research focuses on ultra-small mobile devices that support spatial interaction. Before, he was a research scholar in HCI with Steven Feiner at Columbia University in the City of New York. More information on http://www.christianholz.net/