The new large-scale multi-view Gaussian model (LGM) jointly launched by Peking University, Nanyang Technological University S-Lab and Shanghai Artificial Intelligence Laboratory has made a significant breakthrough in the field of 3D modeling. The core module of this model, LargeMulti-ViewGaussianModel, can generate high-quality 3D object models in just 5 seconds using only single-view images or text. This technology breaks through the bottleneck of efficient 3D representation and high-resolution generation networks, realizes Text-to-3D and Image-to-3D tasks, and provides Mesh conversion methods, providing new opportunities for future 3D modeling technology development. direction.
The large-scale multi-view Gaussian Model (LGM) launched by a team jointly researched by Peking University, Nanyang Technological University S-Lab and Shanghai Artificial Intelligence Laboratory uses the core module LargeMulti-ViewGaussianModel to generate high-quality images or text from a single view within 5 seconds. 3D objects. LGM solves the challenges of efficient 3D representation and generation networks at high resolution, supports Text-to-3D and Image-to-3D tasks, and provides Mesh conversion methods, bringing significant progress.
LGM's fast generation speed and high-quality output have brought revolutionary changes to 3D modeling technology. Its application prospects are broad and it is expected to play an important role in game development, virtual reality, architectural design and other fields. In the future, with the continuous improvement of technology, LGM is expected to further improve efficiency and accuracy, providing users with a more convenient and efficient 3D modeling experience.