TelePresence

Problems with video based telepresence:

- hard to tell who's talking to who in a meeting of 3 or more

- bandwidth needed by continuous video streaming for ongoing telepresence

-- in particular, upload speeds from residential connections

Solution: Gesture Streaming

- 3D avatars are computed and transmitted once

- then facial and limb gestures are tracked and streamed, using less bandwidth

- avatars can be arranged in 3D virtual office, and mouse or head look can point head at specific person

Example 3D virtual meeting, with avatars:

As she speaks Jane can see and notice Joe's skeptical expression -based on squinty micro-gestures around Joe's eyes and slightly puckered mouth micro-gesture - and from that Jane realizes she has some convincing to do: