Gigantic Camera mesh by Google Captures accurate 3D models of People
Google can scan your every data, your every move, your every word, and your body. wait, what? body? yes! Google AI research team has developed a 360 capturing dome that not only captures human body realistically in 3D but can reimagine them in new lighting conditions.
Unlike photo scan model generation as done by quixel (now acquired by Epic Games) which captures multiple photos from every angle and create 3d models from AI algorithm, this technology dubbed as volumetric capture, uses multiple cameras in a dome shaped cage and can not only produces very detailed 3d model including creases of clothes, but can also relight the model in any other situation which can then be placed in a virtual environment.
This uses 330 led lights and 100 12MP cameras capturing at 60Hz which is able to generate light agnostic 3d models.
We expect this demo of the project called “The Relightables” to be presented in Siggraph.
Till then, check out the video of this cool video capture.
- Titan Fall(ter): Intel’s Stumble an Analysis - August 5, 2024
- AMDs Radeon Future Looks Bleak After RDNA 5 - May 1, 2024
- Kinect 2- Right time to use A.I. in Nextgen Consoles? - April 30, 2024
