Register now After registration you will be able to apply for this opportunity online.
Adopting Delaunay tetrahedralization for dynamic NeRFs
Tetra-NeRF [1] offers a way to represent the scene as Delaunay tetrahedralization of the input point cloud. This can be used to represent dynamic 3D scenes [2] as the deformation is performed on the vertices of the tetrahedral mesh.
Keywords: NeRF, 3D scene reconstruction
Tetra-NeRF [1] offers a way to represent the scene as Delaunay tetrahedralization of the input point cloud. This can be used to represent dynamic 3D scenes [2] as the deformation is performed on the vertices of the tetrahedral mesh.
The goal is first to construct a displacement field transforming the vertices of the mesh and train the joint Tetra-NeRF and displacement field representation. Later, the displacement field can be baked into the representation to perform efficient reconstruction.
Another (possible) application is to model physics [3] using the vertices of the tetrahedra.
References:
[1] Kulhanek J., Sattler, T. Tetra-NeRF: Representing Neural Radiance Fields Using Tetrahedra, ICCV 2023
[2] Park, K, et. al. Nerfies: Deformable Neural Radiance Fields, ICCV 2021
[3] Xie, T. PhysGaussian: Physics-Integrated 3D Gaussians for Generative Dynamics.
Tetra-NeRF [1] offers a way to represent the scene as Delaunay tetrahedralization of the input point cloud. This can be used to represent dynamic 3D scenes [2] as the deformation is performed on the vertices of the tetrahedral mesh. The goal is first to construct a displacement field transforming the vertices of the mesh and train the joint Tetra-NeRF and displacement field representation. Later, the displacement field can be baked into the representation to perform efficient reconstruction. Another (possible) application is to model physics [3] using the vertices of the tetrahedra.
References: [1] Kulhanek J., Sattler, T. Tetra-NeRF: Representing Neural Radiance Fields Using Tetrahedra, ICCV 2023 [2] Park, K, et. al. Nerfies: Deformable Neural Radiance Fields, ICCV 2021 [3] Xie, T. PhysGaussian: Physics-Integrated 3D Gaussians for Generative Dynamics.
Using Tetra-NeRF to represent and reconstruct dynamic 3D scenes
Using Tetra-NeRF to represent and reconstruct dynamic 3D scenes