Put simply, Radiance Field methods have revolutionized novel-view synthesis of scenes captured with multiple photos or videos, or generating 3D scenes from 2D images. Unfortunately, achieving high visual quality still requires costly AI-powered neural networks, but this new 3D gaussian splatting technology by researchers from University Cote d’Azur as well as the Max-Planck Institute changes that.
This method allows for high-quality real-time (≥ 100 fps) novel-view synthesis at 1080p resolution. Technically speaking, the researchers start with 3D Gaussians that preserve desirable properties of continuous volumetric radiance fields for scene optimization while avoiding unnecessary computation in empty space. They then develop a fast visibility-aware rendering algorithm that supports anisotropic splatting and both accelerates training as well as enables real-time rendering.
- Experience total immersion with 3D positional audio, hand tracking and easy-to-use controllers working together to make virtual worlds feel real.
- Explore an expanding universe of over 500 titles across gaming, fitness, social/multiplayer and entertainment, including exclusive releases and...
- Enjoy fast, smooth gameplay and immersive graphics as high-speed action unfolds around you with a fast processor and immersive graphics.
Note that for comparable training times to InstantNGP, we achieve similar quality to theirs; while this is the maximum quality they reach, by training for 51min we achieve state-of-the-art quality, even slightly better than Mip-NeRF360,” said the researchers.
