Ray tracing volume densities

By J. Kajiya, B. V. Herzen. 1984

Subfields: Computer Graphics and Computer-Aided Design

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Graphics and Computer-Aided Design

Citations in the corpus

GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fields (2020, background): “The volume rendering operator π vol [37] maps these evaluations to the pixel’s ﬁnal feature vector f : Using numerical integration as in [61], f is obtained as where τ j is the transmittance, α j the alpha value for x j , and δ j = || x j +1 − x j || 2 the distance between neighboring sample points.”
StyleSDF: High-Resolution 3D-Consistent Image and Geometry Generation (2021, methodology): “NeRF introduced the use of volume rendering [32] for reconstructing a 3D scene as a combination of neural radiance and density fields to synthesize novel views.”
Visual simulation of smoke (2001, background): “Kajiya and Von Herzen were the first in CG to do this [13].”
Real-time volume graphics (2006, methodology): “Light Transport The traditional volume rendering equation proposed by Levoy [68] is a simpli.ed approximation of a more general volumetric light transport equation .rst used in computer graphics by Kajiya [44].”
NeRF-Editing: Geometry Editing of Neural Radiance Fields (2022, methodology): “NeRF samples points on the ray and uses volume rendering [30] to obtain the color of each ray.”
Neural Body: Implicit Neural Representations with Structured Latent Codes for Novel View Synthesis of Dynamic Humans (2020, methodology): “The pixel colors are estimated via the volume rendering integral equation [26] that accumulates volume densities and colors along the corresponding camera ray.”
iNeRF: Inverting Neural Radiance Fields for Pose Estimation (2020, methodology): “ong the ray, a set of points are sampled for use as input to the MLP which outputs a set of densities and colors. These values are then used to approximate the image formation behind volume rendering [11] using numerical quadrature [24], producing an estimate of the color of that pixel. NeRF is trai”
Simulation of cloud dynamics on graphics hardware (2003, unknown): “[Kajiya and Von Herzen 1984] used a simple method based on PDEs to generate cloud data sets for their ray tracing algorithm.”
GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis (2020, background): “Neural Radiance Fields: A radiance ﬁeld is a continuous mapping from a 3D location and a 2D viewing direction to an RGB color value [20, 30].”
IBRNet: Learning Multi-View Image-Based Rendering (2021, methodology): “…of a ray r passing through the scene, we ﬁrst query the colors and densities of M samples on the ray, and then perform classic volume rendering [25] with the integral approximated using numerical quadrature [39]: where Here samples from 1 to M are sorted to have ascending depth values. c k and…”
NeRF++: Analyzing and Improving Neural Radiance Fields (2020, methodology): “…to optimize σ and c by minimizing the discrepancy between the ground truth observed images I i , and the predicted images ˆ I i ( σ, c ) rendered from σ and c at the same viewpoints: (1) The implicit volumes σ and c are ray-traced to render each pixel of ˆ I ( σ, c ) (Kajiya & Von Herzen, 1984).”
Depth-supervised NeRF: Fewer Views and Faster Training for Free (2021, methodology): “NeRFs [18] learn an implicit function to model a volumetric radiance field using standard volumetric rendering [8] and alpha compositing techniques [21].”
NeRF: Neural Radiance Field in 3D Vision, A Comprehensive Review (2022, unknown): “In more detail, given volume density and color functions, volume rendering [20] is used to obtain the color C(r) of any camera ray r(t) = o+td, with camera position o and viewing direction d using”