Gravity + Multitouch = 3D tracking

Bootstrapper (note at CHI 2012) recognizes tabletop users by their shoes. Bootstrapper uses a Kinect attached to a Microsoft Surface table pointed at users' shoes.

CapStones and ZebraWidgets (note CHI 2012) are stackable building blocks, dials and widgets for capacitive touch screens that work by handing down capacitance. 

Multitoe (fullpaper UIST 2010) is an interactive floor. Based on FTIR sensing, it can identify users based on their soles, track users' foot and body postures, and enable high-precision interaction-using feet.

Lumino (best paper award CHI 2010, demos at Siggraph 2010 and Tabletop 2010) is a system of building blocks that allows microsoft surface to sense building elements arranged in three-dimensional structures.

RidgePad was our first project to explore 3D reconstruction from contact area.

Imaginary interfaces = spatial gestures

Imaginary Phone (fullpaper UIST 2011) allows users to operate their phone without taking it out of the pocket. Instead users type on their hand. Users learn the device automatically by transferring spatial knowledge from the use of the physical device. 

Data Miming (fullpaper CHI 2011). Users retrieve 3D objects from a database by describing their shape through gesture in 3-space. Tracked using a Kinect/PrimeSense depth camera.

Imaginary Interfaces (fullpaper UIST 2010) are devices that allow users to interact spatially as they normally would with a touch screen--yet without the screen.

Understanding touch

Understanding touch (fullpaper at CHI 2011). In order for a touch device to be highly effective it needs to match users' mental model of touch. But what is that model?

Touch on Curved Surfaces (fullpaper CHI 2011). Technology is emerging that allow touch-enabling non-planar surfaces, e.g., on mobile devices. We are modeling how users interact with such surfaces.

Generalized perceived input point model & ridgepad (fullpaper at CHI 2010). We find that the touch location sensed by capacitive touchpads varies across users and finger postures. We build a device based on a fingerprint scanner that exploits this.

Miniature mobile devices

Rock-Paper-Fibers: (note CHI 2012) Bringing Physical Affordance to Mobile Touch Devices.

Deformable touch devices based on time domain reflectometry (fullpaper UIST 2011) allow touch-enabling stretchable, reconfigurable, and modular devices, all with a single two-wire cable.

Nenya (note at CHI 2011). This input device is a passive magnetic ring. We obtain this tiny device form-factor by offloading all electronics into a bracelet that tracks the ring using a magnetometer.

Touch Projector (fullpaper CHI 2010) allows users to manipulate the contents located on distant public displays by touch manipulating its video image on their mobile device.

Disappearing Mobile Devices (fullpaper UIST 2009) are devices so small, that they only allow certain types of gesture interactions: a thought experiment about the ultimate future of mobile miniaturization.

Nanotouch (best paper nominee CHI 2009) is a prototype device that users operate via the device backside. Allows for making very small mobile devices.

Others

Mulloni, A., Duenser, A., Seichter, H., Baudisch, P., Schmalstieg, D. 360° Panoramic Overviews for Location-Based Services. In Proceedings of CHI 2012, 4 pages (Note). 

My new PC is a mobile phone (keynote given at Mobile HCI 2010 and USAB 2010) I argue that the computational device for the world is already here--it is mobile phones. 4 billion of them. But to make them useful, we need to create new software that turns phones into stand-alone computers.
PPT+Video (116.3M)

Relaxed Selection Techniques (fullpaper UIST 2009) allow users to search time series data using a pen stroke that not only specifies the shape of the sought segment, but also specifies tolerances.