Virtual Humans on Mobile Devices

Project Leader: Ari Shapiro

Complex virtual humans are typically found in museums, specialized training installations, and similar settings. In many cases, a virtual human resides in a particular place, and is often only available under formal or brief encounters. Mobile platforms such as smartphones and tablets are a pervasive technology, and are now capable of running the software and hardware components that are necessary for a convincing, interactive Virtual Human. Potentially, a virtual human running on a mobile platform changes three significant relationships with real humans:
(1) a mobile virtual human can be accessible to a mobile user at any time,
(2) a mobile virtual human can be accessible to a mobile user at any location, and
(3) the group of real users who could potentially interact with a mobile virtual human is broadened to all people with a smartphone.
Thus, rich, long-term interactions with a broad range of types of people would now be possible on mobile devices, through a broad set of domains.

We study the effect of using virtual humans on mobile devices and develop architectures for rapidly building virtual humans and mobile devices that include speech, nonverbal behavior, and interaction.

We present studies of how the presentation of a virtual human on a mobile device can affect the user interaction. For example, does an embodied virtual human elicit better copresence than an audio-only interface? Does using a virtual character with video differ from using a virtual character presented using 3D technologies?