原帖引用 · 原文
Rohan Paul@rohanpaul_ai
Very good news for robot training, since the rarest scenarios can be rerun on demand instead of waited for.
@ZeromatterTech is out of stealth after 4 years spent building a physically accurate copy of the world, where cars, drones, aircraft and robots learn before they move on real roads, in factories or in homes.
They wrote their own renderer, physics engine and simulation platform, built for how cameras, lidar and radar see. The same test gives the same result every time, and results are checked against real data from their own labs.
So the rare crash gets rehearsed a million times in sim, long before the road sees it.
4 years ago, Zeromatter started with a single bet: building autonomous machines would be faster, safer, and better done first in a virtual world.
Today, we’re coming out of stealth to show what that bet has become. https://t.co/HijEZUXcL4
在 X 查看这条帖子 ↗
· UTC+8