The virtual Q version of Huang Renxun appeared at the NVIDIA Developer Conference: We have the ability to create a new 3D world

Huang Renxun said that Omniverse, the Nvidia virtual world construction platform, is a “digital wormhole” that can connect a single Omniverse world to another world.

The virtual Q version of Huang Renxun appeared at the NVIDIA Developer Conference: We have the ability to create a new 3D world

Image source: NVIDIA

On November 10th, NVIDIA , which is dominant in the field of artificial intelligence (AI) computing , released a variety of new software and hardware products at its developer conference, mainly for data centers, quantum computing, and virtual world construction. Among them, Omniverse, the Nvidia virtual world construction platform, has been highlighted.

Omniverse launched the beta version in December last year. It was defined by NVIDIA as a simulation and collaboration platform, based on the Universal Scene Description (USD) of the American animation production company Pixar. It supports multi-person collaboration and work, and can be connected to and invoked by multiple 3D software platforms. , And finally provided by the GPU in real-time rendering, which can be used to design factories, train artificial intelligence, and autopilot. Nvidia CEO Huang Renxun said that Omniverse is currently in the open beta stage and has more than 70,000 users.

Based on Omniverse, Nvidia expands virtual world applications to more subdivided areas, such as Omniverse Replicator, which is a synthetic data generation engine for virtual environment training. Since the construction of the Omniverse virtual environment requires huge costs and a large amount of data, Omniverse Replicator can help developers generate a large amount of data required to train the deep neural network that constitutes artificial intelligence.

Omniverse Avatar uses AI and graphics technology to construct human avatars in the virtual world, catering to the popular digital human craze. Huang Renxun introduced that the virtual image can simulate the tone and pronunciation of a person, and can see, speak, and talk about various topics, as well as a reasonable understanding of the original intention of the questioner and express the intention. According to Huang Renxun’s vision, based on Omniverse Avatar, AI assistants can provide more customized interactive services for different industries, such as restaurant ordering, bank transactions, etc. At the press conference, Huang Renxun displayed the customer service scene based on his own prototype, and created a virtual “Q version of Huang Renxun” that can answer questions smoothly and accurately.

Prior to this, Nvidia went out of the circle because of its “digital double “. In a speech video in April this year, there were about 15 seconds of using the virtual Huang Renxun created by Omniverse and more than 6000 objects in the kitchen, and the others were all real people. In order to achieve these 15 seconds, NVIDIA scanned Huang Renxun’s whole body image to create a 3D model, and then a real person imitated Huang Renxun’s gestures and facial expressions during a speech to achieve motion capture, and used the collected data to train the AI ​​model, and finally combined the two Two into one.

For the application of “digital twin” technology in the virtual industrial field, NVIDIA has made many progress. For example, BMW uses Omniverse to plan future factory design in a virtual environment , including space configuration, production line, human-machine collaboration planning, etc. Physical simulation combined with AR and other technologies can enable humans to experience and find problems in the virtual factory space, making factory planning more complete; Ericsson builds a city in Omniverse as a test field for 5G technology, and uses physical simulation technology to detect 5G The signal transmission in the virtual environment serves as a reference for improving the configuration of communication base stations and optimizing antenna design.

Nvidia believes that Omniverse can be used to train artificial intelligence technologies such as robots and autonomous vehicles. For example, to adapt to the customized needs of today’s customers to create a flexible factory assembly line, you can first simulate the assembly line, workers, and robots through the digital version to find the optimal adjustment or configuration plan; Carry out enough simulation tests on the road.

“You will see an unchanging theme-how Omniverse can be used to simulate warehouses, factories, physical and biological systems, 5G edges, robots, self-driving cars, and even virtual images.” Huang Renxun said, he emphasized that Omniverse and The game engine is different, it is designed on the scale of the data center, and hopes to reach the planetary scale in the future.

It can be said that Omniverse is the final result of NVIDIA’s long-term investment in graphics and AI. The construction of Omniverse needs to consume a lot of computing power to restore the physical world in the virtual world, which requires NVIDIA’s GPU, network technology, AI algorithm, etc.

“We now have the technology to create a new 3D world or simulate our physical world.” Huang Renxun said that he hopes that Omniverse will one day reach the scale of global data, connecting people and computers to Omniverse. He also said that Omniverse is essentially a “digital wormhole ” that can connect a single Omniverse world to another world.

The raging meta-universe trend has further increased the attractiveness of NVIDIA to build a virtual world. After the press conference, Nvidia shares rose 5% before the US stock market on Tuesday.

Posted by:CoinYuppie,Reprinted with attribution to:
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.

Leave a Reply