Read more at source.
Read more at source.
Cosmos can generate realistic video footage of events such as boxes falling from shelves inside a warehouse. This can be used to train robots to recognize and respond to accidents. Users can also fine-tune the models using their own data, enhancing the model's versatility and applicability across different scenarios.
Nvidia also announced a new feature in its existing Isaac robot simulation platform. This feature allows robot builders to generate large amounts of synthetic training data from a small number of examples of a desired task, such as grasping a particular object. This efficient learning approach could revolutionize the way robots are trained.
Alongside Cosmos, Nvidia announced Project Digits, a $3,000 personal AI supercomputer capable of running a large language model of up to 200 billion parameters without the need for cloud services. In addition, Nvidia unveiled its highly anticipated next-generation RTX Blackwell GPUs and new software tools to aid in the creation of AI agents.
Its not about generating creative content, but teaching the AI to understand the physical world. - Jensen Huang, CEO of Nvidia
