Source: Science and Technology Innovation Board Daily Author: Song Ziqiao
① Nvidia emphasizes “AI+” to achieve breakthroughs in film and television production, 3D modeling, and NPC interaction through generative AI; ② Chinese game makers Tencent, NetEase, and media giants Netflix, Pixar, and Disney Animation Studios will all participate in the discussion.
On the afternoon of March 18, local time (early morning of March 19, Beijing time), the world's top AI event, Nvidia GTC 2024, will officially open. Nvidia CEO Huang Renxun will give a keynote speech. OpenAI,$Microsoft (MSFT.US)$,$Meta Platforms (META.US)$Tech giants such as Google and DeepMind will all send representatives to attend the conference, and Nvidia industry chain companies such as Hon Hai and Guangda will also attend.
During the warm-up period of the conference, Nvidia's latest technological advancements in gaming and video have become a hot topic.
The agenda for this conference included 11 game-related presentations, 34 media and entertainment-related presentations, Chinese game manufacturers$TENCENT (00700.HK)$,$NetEase (NTES.US)$, and media giants$Netflix (NFLX.US)$, Pixar,$Disney (DIS.US)$Animation studios and others will all participate in the discussion. The content of Nvidia's keynote speeches related to games, film and television mainly focused on the following aspects:
1. Generative AI-assisted game development and film production
In his keynote speech, Hwang In-hoon may explore how generative AI and path tracking technology can be used to create more realistic virtual characters and worlds. Advances in these technologies have brought new possibilities for game development and video production, making the creation of virtual environments and characters more efficient and realistic.
The presentation may mention how the development of generative AI technology allows NPCs (non-player characters) to have greater complexity, autonomy, and individuality. This is important for enhancing the game's immersive and dynamic storytelling experience.
Other speakers will also introduce how to use NVIDIA RTX Remix and generative AI technology to “refurbish” classic old games, mainly screen upgrades.
2. Wensheng map, Wensheng video model, and other AI applications
Anastasis Germanidis, the chief technology officer and co-founder of Runway, a popular Wensheng video application developer, will personally be on site to introduce the development and commercialization process of Runway Gen-1 and Gen-2 video generation models.
Tencent, on the other hand, will introduce its new multi-billion parameter Wensheng image model. The model has been widely used in commercial fields such as advertising, social media, games, media, etc., and generates 5 million images every day.
Other AI tools to be introduced include music generation AI applications from music technology startup DigitraX.
3. 3D content generation
(1) Create ultra-realistic 3D characters with one click
A new 3D generative AI technology that can automatically generate ultra-realistic digital characters based on images or text prompts may be introduced at the conference.
The head of NetEase's computer vision team will use the “Justice Mobile” mobile game (Justice Mobile) as an example. It is reported that millions of users have used characters automatically created by AI as avatars.
(2) Run NPCs in a 3D scene
The CEO and Chief Technology Officer of startup Convai will be speaking on how to use AI technology to make NPCs talk to players in 3D scenes and how to achieve low latency interactions through different AI models.
4. Create games in the cloud
Game developers from Tencent will share how to develop games in the cloud, how to reshape the 3D art resource creation process using OpenUSD, and how to integrate these tools into game engines and digital content production software to achieve more efficient game creation.
OpenUSD is a high-performance 3D scene description technology. The OpenUSD Alliance (AOUSD) was co-founded by Nvidia, Pixar, Adobe, Apple, and Autodesk to promote the standardization, development, evolution, and development of OpenUSD. Tencent is a member of this alliance.
Overall, Nvidia emphasizes “AI+” to achieve breakthroughs in film and television production, 3D modeling, and NPC interaction through generative AI.
GTC 2024 provided participants with an opportunity to learn more about and explore the possibilities of future game and film production. Three days from now, 1,045 lectures, exchanges, and training sessions of various forms will be staged in turns at the San Jose Convention Center in the US. Can AI application topics such as games, film, and television stand out from them and make people feel refreshed? We can look forward to it.
edit/lambor