Somraj, N and Soundararajan, R (2023) ViP-NeRF: Visibility Prior for Sparse Input Neural Radiance Fields. In: 2023 Special Interest Group on Computer Graphics and Interactive Techniques Conference, SIGGRAPH 2023, 6 August 2023 through 10 August 2023, Los Angeles.
|
PDF
sig_2023_con_pap_2023.pdf - Published Version Download (5MB) | Preview |
Abstract
Neural radiance fields (NeRF) have achieved impressive performances in view synthesis by encoding neural representations of a scene. However, NeRFs require hundreds of images per scene to synthesize photo-realistic novel views. Training them on sparse input views leads to overfitting and incorrect scene depth estimation resulting in artifacts in the rendered novel views. Sparse input NeRFs were recently regularized by providing dense depth estimated from pre-trained networks as supervision, to achieve improved performance over sparse depth constraints. However, we find that such depth priors may be inaccurate due to generalization issues. Instead, we hypothesize that the visibility of pixels in different input views can be more reliably estimated to provide dense supervision. In this regard, we compute a visibility prior through the use of plane sweep volumes, which does not require any pre-training. By regularizing the NeRF training with the visibility prior, we successfully train the NeRF with few input views. We reformulate the NeRF to also directly output the visibility of a 3D point from a given viewpoint to reduce the training time with the visibility constraint. On multiple datasets, our model outperforms the competing sparse input NeRF models including those that use learned priors. The source code for our model can be found on our project page: https://nagabhushansn95.github.io/publications/2023/ViP-NeRF.html. © 2023 ACM.
| Item Type: | Conference Paper |
|---|---|
| Publication: | Proceedings - SIGGRAPH 2023 Conference Papers |
| Publisher: | Association for Computing Machinery, Inc |
| Additional Information: | The copyright for this article belongs to authors. |
| Keywords: | Interactive computer graphics; Rendering (computer graphics), Encodings; Neural rendering; Novel view synthesis; Performance; Plane sweep; Plane sweep volume; Sparse input neural radiance field; Swept volume; View synthesis; Visibility prior, Visibility |
| Department/Centre: | Division of Electrical Sciences > Electrical Communication Engineering |
| Date Deposited: | 22 Jun 2024 07:45 |
| Last Modified: | 22 Jun 2024 07:45 |
| URI: | http://eprints.iisc.ac.in/id/eprint/85554 |
Actions (login required)
 |
View Item |
