Adaptive Incident Radiance Field Sampling and Reconstruction Using Deep Reinforcement Learning
Contribution
Addresses the light-field sampling and reconstruction problem with deep learning techniques and offline datasets.
Proposes a novel R-network that explores the image and direction spaces of the radiance field to effectively filter and reconstruct the incident radiance field.
Presents a novel RL-based Q-network to guide the adaptive rendering process.
Related Work
Image-Space Methods
Light field Reconstruction Methods
Light field Adaptive Sampling Methods
Filtering using DNN
DRL
R ...
Temporal Coherence-based Distributed Ray Tracing of Massive Scenes
Contribution
An efficient temporal coherence-based scheduling algorithm
A domain assignment algorithm
Assign domains to all nodes according to the ray transmission information among domains in the previous frame
A runtime scheduling algorithm
Estimate each domain’s pre-loading prior based on the node’s current situation and information of the previous frame
A new virtual portal structure to record the radiance of rays passing through domains in the previous frame
In the current frame, pred ...
Vectorization for Fast, Analytic, and Differentiable Visibility
Author: Yang Zhou, Lifan Wu, Ravi Ramamoorthi, Ling-Qi Yan
Paper: Link
Video:
Presentation:
IntroductionThere are generally two approaches to render images:
Rasterization: Project scene geometry onto the screen and breaks the geometry into pixels
Fast but prone to aliasing
Ray Tracing: Cast rays into the scene and bounces them stochastically to find paths connecting the light and the camera
High-quality but is slow and noisy
Main ProblemMain problem for both approaches: point sampling
...
Physics Based Differentiable Rendering: Edge Sampling
IntroductionEdge sampling is a method to calculate the derivative of the ray tracing result w.r.t. some scene parameters (including camera pose, scene geometry, material and light parameters).
The key idea of edge sampling is dividing the gradient integral into smooth (interior) and discontinuous (boundary) regions. For the smooth part, we use automatic differentiation. For the discontinuous, we use edge sampling to capture the changes at boundaries.
Condition:
Focus on triangle meshes
Assume ...