NerfAcc: Efficient Sampling Accelerates NeRFs
read the original abstract
Optimizing and rendering Neural Radiance Fields is computationally expensive due to the vast number of samples required by volume rendering. Recent works have included alternative sampling approaches to help accelerate their methods, however, they are often not the focus of the work. In this paper, we investigate and compare multiple sampling approaches and demonstrate that improved sampling is generally applicable across NeRF variants under an unified concept of transmittance estimator. To facilitate future experiments, we develop NerfAcc, a Python toolbox that provides flexible APIs for incorporating advanced sampling methods into NeRF related methods. We demonstrate its flexibility by showing that it can reduce the training time of several recent NeRF methods by 1.5x to 20x with minimal modifications to the existing codebase. Additionally, highly customized NeRFs, such as Instant-NGP, can be implemented in native PyTorch using NerfAcc.
This paper has not been read by Pith yet.
Forward citations
Cited by 4 Pith papers
-
Efficient Compression of Structured and Unstructured Volumes via Learned 3D Gaussian Representation
An explicit model using learned 3D Gaussians for volume compression encodes geometry explicitly and outperforms implicit neural representations on unstructured volumes with faster training.
-
BulletGen: Improving 4D Reconstruction with Bullet-Time Generation
BulletGen enhances 4D dynamic scene reconstruction from monocular videos by supervising Gaussian optimization with diffusion-generated frames aligned at a bullet-time step, achieving SOTA on novel-view synthesis and tracking.
-
NeuMesh++: Towards Versatile and Efficient Volumetric Editing with Disentangled Neural Mesh-based Implicit Field
A disentangled mesh-vertex neural radiance field enables mesh-guided geometry edits, texture swap/fill/paint, and semantic-guided edits with claimed efficiency and quality gains.
-
An Active Perception Game for Robust Exploration
Develops a game-theoretic estimator for true information gain in active perception that achieves sub-linear regret and shows average gains of 7% information gain and 42% error reduction across simulated and real robot...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.