Neural 3D Reconstruction and Radiance Fields

8 repos

Deep learning approaches for reconstructing 3D scenes and objects from 2D images, including neural radiance fields (NeRF), Gaussian splatting, and volumetric rendering techniques. The cluster focuses on both the theoretical methods and practical systems for scene capture, novel view synthesis, and geometric reconstruction, with heavy use of Python for implementation alongside some C/C++ optimization layers. Repositories here span from core algorithmic work to end-to-end pipelines and datasets that support training and evaluation of these models.

Related clusters