- Core technical competencies
- Enterprise-level outreach: from laboratory to conference room
- Practical application of scenes and visions
Google announced at its I/O 2025 General Assembly that Project Starline, which existed as a research project, had formally evolved into a new 3D video communication platform - Google Beam. Building on Google Cloud and advanced AI technology, the platform works towards a more natural, immersed video talk experience.
What does it do?Google Beam is a 3D effect that allows remote video calls to look like real face-to-face chatting, using artificial intelligence to capture the 2D image “Change” found on a regular camera.
-
** What are the main functions?** 3D video call (no need to wear equipment)
-
Real-time voice translation (retention of tone, expression)
-
Natural eye contact and expression.
-
Compatibility with enterprise-level workflows
-
** What problem does it solve?** The “dissociation” and language barriers of cross-geographic communication make remote collaboration more like being physically present.
-
** What’s it using?** Build a volume video model through AI + optical display technology.
-
** What has been achieved?** A immersion, high level of integrity and humanization of remote communication has been achieved, especially in the context of business teleconferencing, client exchanges, etc.
Initially,Project Starline was a front-line research project launched by Google a few years ago to create an experience that allows remote video exchanges to be as accessible as they are. Its core objective is:** to break physical space limits and achieve a sense of face-to-face communication** and not to rely on any additional equipment, such as VR helmets or 3D glasses. Google now announces the full upgrade of the project, named Google Beam - a 3D video communication platform driven by ** artificial intelligence**, marking an important turning point in its transition from laboratory research to commercial deployment.
Core technical competencies
AI-driven “Volumetage Video Model” (Volumetric Video Model) The core of Google Beam is an advanced AI model capable of automatically reconstructing 3D images with depth and stereo sense from a normal 2D video stream. This means:
-
Users can see each other in real proportions, angles, light without wearing any equipment, as if there was a person sitting across the street.
-
Support for a more natural and flexible interaction from an arbitrary perspective.
** The Light Field Display displays a real sense of space** In conjunction with the AI model output video, Beam creates a very real depth by displaying technology in the field:
-
Real eye communication**, unlike the traditional video, which is “not in line”;
-
Capable of capturing and transmitting subtle** facial expressions and physical movements**, which contribute to understanding, building trust and resonance.
- AI Real-time voice translation: a natural bridge for cross-linguistic communication Google Beam is integrating ** real-time voice translation** and is currently applied in Google Meet:
-
Support for “almost real-time” translations;
-
** Retain original tone, tone and emotional expression**;
-
Allowing users of different languages to speak the same language,** maximizing a return to the true feelings of face-to-face dialogue**
This is a crucial step towards realizing Google’s vision of “global accessibility communication”.
Enterprise-level outreach: from laboratory to conference room
Google is working with a number of well-known businesses to introduce Google Beam to practical use:
- ** Device partners: HP Initial roll-out of Beam hardware equipment**
-
HP will show the first Google Beam devices at the InfoComm exhibition in 2025;
-
It will be open to selected business clients during the year.
- Technology integration and distribution network
-
Google, in collaboration with Zoom, plans to integrate Beam functions into the Zoom platform;
-
Working with AVI-SPL, a leading global audio-visual integrator, to promote Beam’s landing among corporate clients.
- ** Lead client pilot enterprise** Among them are the following globally renowned institutions that have been involved in piloting or expressing a willingness to cooperate:
-
Deloitte considers Beam to be “the next generation of technology that connects humanity”, not only as a tool, but also as a new concept of connectivity;
-
Salesforce, Citadel, NEC, Duolingo, Recruit have been deployed internally or have expressed a high level of interest;
-
Clients in the medical sector such as Hackensack Meridian Health are also exploring their potential in telemedicine.
Practical application of scenes and visions
Google Beam was originally designed to “** achieve the same communication experience as one from anywhere**”. Its future application orientation includes, but is not limited to:
-
Teleconferencing, client visits, staff training;
-
Quality interaction between teachers and students in educational scenes;
-
In-depth communication between doctors and remote patients in the medical profession;
-
It may even extend to online recruitment, cultural exchanges, cross-border business negotiations, etc.
Read more on beam.google