Learning Neural Light Fields with Ray-Space Embedding Networks

12/02/2021
by   Benjamin Attal, et al.
0

Neural radiance fields (NeRFs) produce state-of-the-art view synthesis results. However, they are slow to render, requiring hundreds of network evaluations per pixel to approximate a volume rendering integral. Baking NeRFs into explicit data structures enables efficient rendering, but results in a large increase in memory footprint and, in many cases, a quality reduction. In this paper, we propose a novel neural light field representation that, in contrast, is compact and directly predicts integrated radiance along rays. Our method supports rendering with a single network evaluation per pixel for small baseline light field datasets and can also be applied to larger baselines with only a few evaluations per pixel. At the core of our approach is a ray-space embedding network that maps the 4D ray-space manifold into an intermediate, interpolable latent space. Our method achieves state-of-the-art quality on dense forward-facing datasets such as the Stanford Light Field dataset. In addition, for forward-facing scenes with sparser inputs we achieve results that are competitive with NeRF-based approaches in terms of quality while providing a better speed/quality/memory trade-off with far fewer network evaluations.

READ FULL TEXT

page 7

page 8

page 12

research
07/24/2022

Learning Generalizable Light Field Networks from Few Images

We explore a new strategy for few-shot novel view synthesis based on a n...
research
05/29/2023

Volume Feature Rendering for Fast Neural Radiance Field Reconstruction

Neural radiance fields (NeRFs) are able to synthesize realistic novel vi...
research
08/22/2023

Efficient View Synthesis with Neural Radiance Distribution Field

Recent work on Neural Radiance Fields (NeRF) has demonstrated significan...
research
03/31/2022

R2L: Distilling Neural Radiance Field to Neural Light Field for Efficient Novel View Synthesis

Recent research explosion on Neural Radiance Field (NeRF) shows the enco...
research
04/01/2022

LASER: LAtent SpacE Rendering for 2D Visual Localization

We present LASER, an image-based Monte Carlo Localization (MCL) framewor...
research
07/21/2022

AdaNeRF: Adaptive Sampling for Real-time Rendering of Neural Radiance Fields

Novel view synthesis has recently been revolutionized by learning neural...
research
08/29/2023

Efficient Ray Sampling for Radiance Fields Reconstruction

Accelerating neural radiance fields training is of substantial practical...

Please sign up or login with your details

Forgot password? Click here to reset