NVIDIA significantly updates virtual world building tools for robot development and releases visual language model (VLM) to allow robots to recognize the world
NVIDIA has developed
NVIDIA Opens Portals to World of Robotics With New Omniverse Libraries, Cosmos Physical AI Models and AI Computing Infrastructure | NVIDIA Newsroom
https://nvidianews.nvidia.com/news/nvidia-opens-portals-to-world-of-robotics-with-new-omniverse-libraries-cosmos-physical-ai-models-and-ai-computing-infrastructure
Developers Build Fast and Reliable Robot Simulations with NVIDIA Omniverse Libraries | NVIDIA Technical Blog
https://developer.nvidia.com/blog/developers-build-fast-and-reliable-robot-simulations-with-nvidia-omniverse-libraries/
Omniverse is a platform that allows developers to simulate robots and self-driving cars in 3D space using physics calculations. This latest update introduces ' 3D Gaussian Splitting ' as a 3D space construction method, allowing developers to create virtual worlds using photos of rooms they have taken.
In addition, the robot learning frameworks '
Announcing General Availability for NVIDIA Isaac Sim 5.0 and NVIDIA Isaac Lab 2.2 | NVIDIA Technical Blog
https://developer.nvidia.com/blog/isaac-sim-and-isaac-lab-are-now-available-for-early-developer-preview/
Additionally, Cosmos has released Cosmos Reason, a visual language model (VLM) optimized for physics AI and robotics development. Cosmos Reason is a 7 billion parameter VLM that can analyze video and infer real-world events, such as which employee carried a package.
Cosmos Reason is already being used by NVIDIA's robotics and self-driving car research teams, and companies like Uber are also using it for research and development. You can see how Cosmos Reason works on the demo site below.
cosmos-reason1-7b Model by NVIDIA | NVIDIA NIM
https://build.nvidia.com/nvidia/cosmos-reason1-7b
