
Image via Qian et al / Google Research
Google has just unveiled ChatDirector, a cutting-edge Remote Video Conferencing System (RVCS) designed to blend the best of virtual and in-person meeting experiences. This new technology promises to make remote interactions feel more natural by incorporating advanced features like space-aware visual presence and speech-aware attention transition assistance.
ChatDirector works by transforming participants’ RGB video streams into 3D portrait avatars, which are then rendered in a virtual 3D scene. The innovative setup addresses the common drawbacks of traditional video conferencing systems by providing intuitive visual cues and establishing clear spatial relationships among remote attendees, resulting in fewer speech interruptions and higher levels of user engagement.
Video via Qian et al / Google Research
Not just churning out animated 3D cartoon avatars, Google’s ChatDirector creates realistic 3D models of users’ faces. The mini-mes are then attached to character bodies and positioned within virtual environments. This approach allows for more authentic interactions, as the avatars maintain the user’s facial expressions and simulate eye contact with the person speaking.

Image via Qian et al / Google Research
Users can place their avatars in a variety of settings, from standard office meeting rooms to Medieval castles. They can also adjust the camera angle within the 3D environment, adding a dynamic element to the virtual meeting experience.
Although the technology is still in its early stages and some imagery might appear choppy, the potential of ChatDirector is significant. It hints at a future where 3D avatars and virtual settings could become standard in video conferencing.
[via Mashable and Global Village Space, images via Qian et al / Google Research]