Surface-Aligned Neural Radiance Fields for Controllable 3D Human Synthesis
This work addresses the challenge of creating detailed and controllable 3D human avatars for applications in animation and virtual reality, representing an incremental improvement over prior techniques.
The paper tackles the problem of reconstructing controllable 3D human models from sparse multi-view videos by proposing a surface-aligned neural radiance field method, which achieves higher quality in novel-view and novel-pose synthesis compared to existing methods on datasets like ZJU-MoCap and Human3.6M.
We propose a new method for reconstructing controllable implicit 3D human models from sparse multi-view RGB videos. Our method defines the neural scene representation on the mesh surface points and signed distances from the surface of a human body mesh. We identify an indistinguishability issue that arises when a point in 3D space is mapped to its nearest surface point on a mesh for learning surface-aligned neural scene representation. To address this issue, we propose projecting a point onto a mesh surface using a barycentric interpolation with modified vertex normals. Experiments with the ZJU-MoCap and Human3.6M datasets show that our approach achieves a higher quality in a novel-view and novel-pose synthesis than existing methods. We also demonstrate that our method easily supports the control of body shape and clothes. Project page: https://pfnet-research.github.io/surface-aligned-nerf/.