nerfies and hypernerf
HyperNeRF extends Nerfies by using higher-dimensional latent codes to handle topological changes that deformable NeRFs cannot represent, making them complementary approaches rather than competitors.
About nerfies
google/nerfies
This is the code for Deformable Neural Radiance Fields, a.k.a. Nerfies.
This project helps create realistic 3D models from ordinary video footage of moving subjects, capturing subtle details like changing facial expressions or clothing wrinkles. By feeding it a video and camera information, you get an interactive 3D scene that faithfully reproduces light and deformation. It's designed for researchers, artists, or anyone who needs to generate high-quality, deformable 3D representations of real-world objects or people from video.
About hypernerf
google/hypernerf
Code for "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields".
This project helps researchers and computer graphics professionals generate realistic 3D video scenes from 2D video footage, even when objects in the scene change shape or topology. It takes in standard video files and outputs a 3D model that can be rendered into new videos from different viewpoints. This is ideal for those working on advanced visual effects, virtual reality content, or scientific visualization.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work