MearcAll papers

arXiv 2003.08934 · cs.CV cs.GR

NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis

Ben Mildenhall, Pratul P. Srinivasan, Matthew Tancik, Jonathan T. Barron, Ravi Ramamoorthi, Ren Ng

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.

No reproduction recordedres_02b199b9
In 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)

Introduction · page 1 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_a1c37c34
We 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)

Introduction · page 2 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_6400fd04
NeRF: 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)

Introduction · page 3 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_6d78b2af
We 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)

Introduction · page 3 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_0dd8fa87
As 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)

Introduction · page 3 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_051c78ad
Our 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.

Volume Rendering with Radiance Fields · page 6 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_b430aeaa
In 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.

Optimizing a Neural Radiance Field · page 7 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_dcd4d331
However, we observe that these components are not sufficient for achieving state-of-the-art quality, as demonstrated in Section 6.4).

Optimizing a Neural Radiance Field · page 7 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_52b5affb
We introduce two improvements to enable representing high-resolution complex scenes.

Optimizing a Neural Radiance Field · page 7 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_b9ea4fd2
We 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)

Positional encoding · page 7 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_6c22473c
We 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).

Results · page 9 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_edad93a1
We 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.

Results · page 9 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_675134f2
We report PSNR/SSIM (higher is better) and LPIPS [50] (lower is better).

Datasets · page 10 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_4d09def7
We thoroughly outperform both baselines that also optimize a separate network per scene (NV and SRN) in all scenarios.

Discussion · page 13 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_024b661d
Note 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).

Ablation studies · page 13 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_53967751
Only using 5 frequencies reduces performance, but increasing the number of frequencies from 10 to 15 does not improve performance.

Ablation studies · page 14 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_aac02949
We 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).

Ablation studies · page 14 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_874551ba
We 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)

Conclusion · page 14 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_22eaf5c3
Training 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)

Additional Implementation Details · page 17 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_bebf2e28
The per-scene breakdown is consistent with the aggregate quantitative metrics presented in the paper, where our method quantitatively outperforms all baselines.

Additional Results · page 22 · Automated extraction, not reviewed · 0 reviewed reproductions

Reproduced one of these claims?

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.