DONeRF: Towards Real-Time Rendering of Compact Neural Radiance Fields using Depth Oracle Networks

by   Thomas Neff, et al.

The recent research explosion around implicit neural representations, such as NeRF, shows that there is immense potential for implicitly storing high-quality scene and lighting information in compact neural networks. However, one major limitation preventing the use of NeRF in real-time rendering applications is the prohibitive computational cost of excessive network evaluations along each view ray, requiring dozens of petaFLOPS. In this work, we bring compact neural representations closer to practical rendering of synthetic content in real-time applications, such as games and virtual reality. We show that the number of samples required for each view ray can be significantly reduced when samples are placed around surfaces in the scene without compromising image quality. To this end, we propose a depth oracle network that predicts ray sample locations for each view ray with a single network evaluation. We show that using a classification network around logarithmically discretized and spherically warped depth values is essential to encode surface locations rather than directly estimating depth. The combination of these techniques leads to DONeRF, our compact dual network design with a depth oracle network as its first step and a locally sampled shading network for ray accumulation. With DONeRF, we reduce the inference costs by up to 48x compared to NeRF when conditioning on available ground truth depth information. Compared to concurrent acceleration methods for raymarching-based neural representations, DONeRF does not require additional memory for explicit caching or acceleration structures, and can render interactively (20 frames per second) on a single GPU.


page 1

page 3

page 5

page 6

page 8

page 9

page 11

page 12


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

Novel view synthesis has recently been revolutionized by learning neural...

Light Field Networks: Neural Scene Representations with Single-Evaluation Rendering

Inferring representations of 3D scenes from 2D observations is a fundame...

Efficient View Synthesis with Neural Radiance Distribution Field

Recent work on Neural Radiance Fields (NeRF) has demonstrated significan...

An Attempt of Adaptive Heightfield Rendering with Complex Interpolants Using Ray Casting

In this technical report, we document our attempt to visualize adaptive ...

Rig-space Neural Rendering

Movie productions use high resolution 3d characters with complex proprie...

Towards a Robust Framework for NeRF Evaluation

Neural Radiance Field (NeRF) research has attracted significant attentio...

Fristograms: Revealing and Exploiting Light Field Internals

In recent years, light field (LF) capture and processing has become an i...

Please sign up or login with your details

Forgot password? Click here to reset