Last year, the face of aimed, the AIV platform, released, collection of models open, data set, and tools to help build real robotic systems. On Tuesday, hugging the faces with AI Startup Yeak to enlarge the training with training training for robots and cars that navigate the environment, like the city of the city, with an otonom.
New aircraft, called study for drive (l2d)Through Petabyte size, and contains data from the cars on the car in the German driving school. L2D takes data, GPS, and “Vehicle Dynamics” of driving instructors and students navigating streets with construction zones, intersection, highways, and more.
There are several training training that run out of the company including Waymo and Koma Ai. But many focus on planning plans like object detection and tracking, requiring the highest captivity, according to the L2D creator – making it difficult to scale.
Different, L2D designed to encourage the development of “exotic-resistant”, the authorities, which helps predict the action (eg when walking on the road (eg camera snapshots)
“AI Community can now build an independent driving model,” Harsimry co -der and remi, a AI member for the robotic team, writing in blog posts. “L2D is intended as the largest of the largest driving data of a AI community with an unique and variety of spatial episode episode for the final spatial intelligence.”
Roso Hugging and the Youak Plan to perform the “closed-based model test” covered using L2D and the lerobot summer, which is sent in the vehicle with safety driver. The company calls in AI community to send model and duty to be evaluated, like a circle navigation and parking space.
