Portaltic.-Google details the operation of its 3D communication system based on telepresence

02-12-2021 Starline project concept GOOGLE RESEARCH RESEARCH AND TECHNOLOGY POLICY

MADRID, 2 (Portaltic/EP)

Google researchers have shared the advances they have made with the Starline project, a three-dimensional telepresence communication system, which still has to solve challenges in image capture, projection and the sense of co-presence.

Google announced the Starline project at the last Google I / O developer conference, a communication system that is committed to telepresence, that is, that the two participants in a conversation feel that they are next to each other, physically, despite the distance thanks to 3D technology.

This system is still under development, as referred to in the document published by Google Research, in which they point out that they still have to solve three problems: capturing and rendering the 3D image of the person remotely in a realistic way, creating a high-resolution projection with making the user feel comfortable and achieving co-presence, which includes proximity, eye contact and interaction.

“Telepresence presents enormous opportunities to bring together organizations and social groups that are increasingly distributed throughout the world,” say those responsible for this project. Compared to 2D systems, those who have participated in the studies have noted that the 3D system is “significantly better at fostering various elements of communication, including presence, attention, reaction measurement, eye contact, engagement, and communication. the personal connection “.

The researchers also indicate that the participants “remembered more meetings and demonstrated more non-verbal behaviors (hand gestures, nods of the head, and eyebrow movements) than in the 2D videoconferences,” as reported in the article.

The system they have developed uses a Lenovo P920 computer and Nvidia graphics cards (two Quadro RTX 6000 and two Titan RTX) for 60Hz video processing (which rises to 120Hz in the case of face monitoring).

It also consists of two structures: a screen, with cameras, loudspeakers, illuminating microphones, and the computer, where the contact would be represented, and another rear unit with infrared feedback in which there is a bench for the user to sit.

“Although it reaches a level of audiovisual fidelity not demonstrated in previous telepresence systems, some effects are still not captured well” in which they are working, recognize the researchers, who point out as an example that elements such as hair or glasses can give rise to geometry and texture errors.

They also have to work on depth, stereo sound, color and comprehension systems, as well as pixel density and panel architecture to make conversations with this system realistic, and to expand the number of participants and extend this mode of communication to groups.



Reference-www.infobae.com

Leave a Reply

Your email address will not be published. Required fields are marked *