Lagrangian Hashing for Compressed Neural Field Representations
Preprint
Abstract
We present Lagrangian Hashing, a representation for neural fields combining the characteristics of fast training NeRF methods that rely on Eulerian grids (i.e.~InstantNGP), with those that employ points equipped with features as a way to represent information (e.g. 3D Gaussian Splatting or PointNeRF). We achieve this by incorporating a point-based representation into the high-resolution layers of the hierarchical hash tables of an InstantNGP representation. As our points are equipped with a field of influence, our representation can be interpreted as a mixture of Gaussians stored within the hash table. We propose a loss that encourages the movement of our Gaussians towards regions that require more representation budget to be sufficiently well represented. Our main finding is that our representation allows the reconstruction of signals using a more compact representation without compromising quality.Keywords
All Related Versions
This publication has 30 references indexed in Scilit:
- PointNeRF++: A Multi-scale, Point-Based Neural Radiance Field2024
- TensoRF: Tensorial Radiance Fields2022
- Neural Point-Based Graphics2020
- NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis2020
- Deep Implicit Volume Compression2020
- Image Compression with Anisotropic DiffusionJournal of Mathematical Imaging and Vision, 2008
- Point-Based Graphics2007
- A survey of point-based techniques in computer graphicsComputers & Graphics, 2004
- Adaptive background mixture models for real-time tracking2003
- Theoretical foundations of transform codingIEEE Signal Processing Magazine, 2001