Posted 3/24/2024, 5:00:08 AM
Nvidia's New AI Model Generates High-Quality 3D Models from Text in Under a Second
- Nvidia unveiled LATTE3D, a new AI model that generates 3D models from text prompts in under a second
- The quality of the 3D models is high and the speed is an order of magnitude faster than previous models
- LATTE3D was trained using prompts from ChatGPT to handle various ways of describing objects
- The model can optimize and export the 3D models for use in graphics software and platforms like Omniverse
- 3D generation is the next frontier for AI after images and video, with rapid recent advances in speed and quality