res_02b199b9In this work, we address the long-standing problem of view synthesis in a new way by directly optimizing parameters of a continuous 5D scene representation to minimize the error of … (excerpt)
arXiv 2003.08934 · cs.CV cs.GR
Read on arXiv · arXiv non-exclusive license
Claims were extracted automatically by reducer-1 on 2026-11-12. The paper's license does not allow republishing it, so long claims are shortened to a quotation. Read the full text on arXiv.
res_02b199b9In this work, we address the long-standing problem of view synthesis in a new way by directly optimizing parameters of a continuous 5D scene representation to minimize the error of … (excerpt)
res_a1c37c34We address these issues by transforming input 5D coordinates with a positional encoding that enables the MLP to represent higher frequency functions, and we propose a hierarchical sampling procedure to … (excerpt)
res_6400fd04NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis 3 – A positional encoding to map each input 5D coordinate into a higher dimensional space, which enables us to … (excerpt)
res_6d78b2afWe demonstrate that our resulting neural radiance field method quantitatively and qualitatively outperforms state-of-the-art view synthesis methods, including works that fit neural 3D representations to scenes as well as works … (excerpt)
clm_0dd8fa87As far as we know, this paper presents the first continuous neural scene representation that is able to render high-resolution photorealistic novel views of real objects and scenes from RGB … (excerpt)
clm_051c78adOur method predicts the changing specular appearance of these two 3D points, and in (c) we show how this behavior generalizes continuously across the whole hemisphere of viewing directions.
clm_b430aeaaIn the previous section we have described the core components necessary for modeling a scene as a neural radiance field and rendering novel views from this representation.
res_dcd4d331However, we observe that these components are not sufficient for achieving state-of-the-art quality, as demonstrated in Section 6.4).
clm_52b5affbWe introduce two improvements to enable representing high-resolution complex scenes.
res_b9ea4fd2We leverage these findings in the context of neural scene representations, and show that reformulatingFΘ as a composition of two functions FΘ =F′Θ◦γ, one learned and one not, significantly improves … (excerpt)
res_6c22473cWe quantitatively (Tables 1) and qualitatively (Figs. 8 and 6) show that our method outperforms prior work, and provide extensive ablation studies to validate our design choices (Table 2).
clm_edad93a1We urge the reader to view our supplementary video to better appreciate our method’s significant improvement over baseline methods when rendering smooth paths of novel views.
clm_675134f2We report PSNR/SSIM (higher is better) and LPIPS [50] (lower is better).
clm_4d09def7We thoroughly outperform both baselines that also optimize a separate network per scene (NV and SRN) in all scenarios.
res_024b661dNote that our method’s performance using only 25 input images still exceeds NV, SRN, and LLFF across all metrics when they are provided with 100 images (see supplementary material).
res_53967751Only using 5 frequencies reduces performance, but increasing the number of frequencies from 10 to 15 does not improve performance.
res_aac02949We believe the benefit of increasing L is limited once 2 L exceeds the maximum frequency present in the sampled input images (roughly 1024 in our data).
clm_874551baWe demonstrate that representing scenes as 5D neural radiance fields (an MLP that outputs volume density and view-dependent emitted radiance as a function of 3D location and 2D viewing direction) … (excerpt)
clm_22eaf5c3Training Details For real scene data, we regularize our network by adding random Gaussian noise with zero mean and unit variance to the output σ values (before passing them through … (excerpt)
clm_bebf2e28The per-scene breakdown is consistent with the aggregate quantitative metrics presented in the paper, where our method quantitatively outperforms all baselines.
Record your result in the Mearc app on your own machine, then use Publish this result with this paper's id (2003.08934) and the claim id shown next to the claim. Mearc creates a file that holds only what you choose to publish. Add it to the public claim data by pull request; a maintainer other than you reviews it before it counts.