2000 character limit reached
Alpha Invariance: On Inverse Scaling Between Distance and Volume Density in Neural Radiance Fields (2404.02155v2)
Published 2 Apr 2024 in cs.CV
Abstract: Scale-ambiguity in 3D scene dimensions leads to magnitude-ambiguity of volumetric densities in neural radiance fields, i.e., the densities double when scene size is halved, and vice versa. We call this property alpha invariance. For NeRFs to better maintain alpha invariance, we recommend 1) parameterizing both distance and volume densities in log space, and 2) a discretization-agnostic initialization strategy to guarantee high ray transmittance. We revisit a few popular radiance field models and find that these systems use various heuristics to deal with issues arising from scene scaling. We test their behaviors and show our recipe to be more robust.
- Mip-nerf: A multiscale representation for anti-aliasing neural radiance fields. In Int. Conf. Comput. Vis., 2021.
- Mip-NeRF 360: Unbounded anti-aliased neural radiance fields. In IEEE Conf. Comput. Vis. Pattern Recog., 2022.
- Zip-NeRF: Anti-aliased grid-based neural radiance fields. In Int. Conf. Comput. Vis., 2023.
- Christopher M Bishop. Neural networks for pattern recognition. Oxford university press, 1995.
- Hexplane: A fast representation for dynamic scenes. In IEEE Conf. Comput. Vis. Pattern Recog., 2023.
- Tensorf: Tensorial radiance fields. In Eur. Conf. Comput. Vis., 2022.
- Raanan Fattal. Single image dehazing. ACM Trans. Graph., 2008.
- Plenoxels: Radiance fields without neural networks. In IEEE Conf. Comput. Vis. Pattern Recog., 2022.
- Single image haze removal using dark channel prior. IEEE Trans. Pattern Anal. Mach. Intell., 2010.
- Delving deep into rectifiers: Surpassing human-level performance on ImageNet classification, 2015.
- Normalization techniques in training dnns: Methodology, analysis and application. IEEE Trans. Pattern Anal. Mach. Intell., 2023.
- 3D Gaussian splatting for real-time radiance field rendering. ACM Trans. Graph., 2023.
- Approximate differentiable rendering with algebraic surfaces. In Eur. Conf. Comput. Vis., 2022.
- Flexible techniques for differentiable rendering with 3d gaussians. arXiv preprint arXiv:2308.14737, 2023.
- Dynamic 3d Gaussians: Tracking by persistent dynamic view synthesis. arXiv preprint arXiv:2308.09713, 2023.
- NeRF in the wild: Neural radiance fields for unconstrained photo collections. In IEEE Conf. Comput. Vis. Pattern Recog., 2021.
- Nelson Max. Optical models for direct volume rendering. IEEE Trans. Vis. Comput. Graph., 1995.
- Progressively optimized local radiance fields for robust view synthesis. In IEEE Conf. Comput. Vis. Pattern Recog., 2023.
- Local light field fusion: Practical view synthesis with prescriptive sampling guidelines. ACM Trans. Graph., 2019.
- Nerf: Representing scenes as neural radiance fields for view synthesis. In Eur. Conf. Comput. Vis., pages 405–421, 2020.
- Instant neural graphics primitives with a multiresolution hash encoding. ACM Trans. Graph., 2022.
- Chromatic framework for vision in bad weather. In IEEE Conf. Comput. Vis. Pattern Recog., 2000.
- Vision in bad weather. In Int. Conf. Comput. Vis., 1999.
- Differentiable volumetric rendering: Learning implicit 3d representations without 3d supervision. In IEEE Conf. Comput. Vis. Pattern Recog., 2020.
- Soft 3d reconstruction for view synthesis. ACM Trans. Graph., 2017.
- Compositing digital images. In Proceedings of Computer graphics and interactive techniques, 1984.
- D-NeRF: Neural radiance fields for dynamic scenes. In IEEE Conf. Comput. Vis. Pattern Recog., 2021.
- MERF: Memory-efficient radiance fields for real-time view synthesis in unbounded scenes. ACM Trans. Graph., 2023.
- Pushing the boundaries of view extrapolation with multiplane images. In IEEE Conf. Comput. Vis. Pattern Recog., 2019.
- Direct voxel grid optimization: Super-fast convergence for radiance fields reconstruction. In IEEE Conf. Comput. Vis. Pattern Recog., 2022.
- Nerfstudio: A modular framework for neural radiance field development. arXiv preprint arXiv:2302.04264, 2023.
- Jiaxiang Tang. Torch-ngp: a PyTorch implementation of instant-ngp, 2022. https://github.com/ashawkey/torch-ngp.
- Neus: Learning neural implicit surfaces by volume rendering for multi-view reconstruction. In Adv. Neural Inform. Process. Syst., 2021.
- Wikipedia. Rectified Gaussian distribution — Wikipedia, the free encyclopedia. http://en.wikipedia.org/w/index.php?title=Rectified%20Gaussian%20distribution&oldid=1140140064, 2023.
- 4d Gaussian splatting for real-time dynamic scene rendering. arXiv preprint arXiv:2310.08528, 2023.
- Multiview neural surface reconstruction by disentangling geometry and appearance. In Adv. Neural Inform. Process. Syst., 2020.
- Volume rendering of neural implicit surfaces. In Adv. Neural Inform. Process. Syst., 2021.
- Lin Yen-Chen. NeRF-Pytorch. https://github.com/yenchenlin/nerf-pytorch/, 2020.
- Stereo magnification: Learning view synthesis using multiplane images. In SIGGRAPH, 2018.
- High-quality video view interpolation using a layered representation. ACM Trans. Graph., 2004.
Collections
Sign up for free to add this paper to one or more collections.