Stochastic Approach to realistic rendering in computer graphics for Virtual Reality (VR)

In Virtual Reality, it is important to make the rendering of images displayed on an HMD headset as realistic as possible. In this case, it is a question of taking into account the multiple reflections, refractions and diffusions using stochastic methods (Monte Carlo, Gaussian processes, etc.) or AI...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2023 International Conference on Networking and Advanced Systems (ICNAS) s. 1
Hlavní autor: Bouatouch, Kadi
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 21.10.2023
Témata:
Geometric modeling
Headphones
Image color analysis
Monte Carlo methods
Neural networks
Resists
Virtual reality
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:In Virtual Reality, it is important to make the rendering of images displayed on an HMD headset as realistic as possible. In this case, it is a question of taking into account the multiple reflections, refractions and diffusions using stochastic methods (Monte Carlo, Gaussian processes, etc.) or AI (CNN neural networks). These methods can be used in other fields such as vision, crowd simulation, etc. This Keynote Lecture begins by introducing the different processing necessary for obtaining a computer-generated image such as: geometric modeling, camera parameters, color, photometric quantities, etc. Then it presents in detail the various stochastic rendering methods allowing the generation of synthetic images with a high level of realism. Among these methods we can cite: the Monte Carlo method, Bayesian Monte Carlo, Metropolis, spectral Analysis of Quadrature Rules and Fourier Truncation-based Methods Applied to the Shading Integral, etc. This lecture will be illustrated with realistic images generated with different rendering methods.
DOI:10.1109/ICNAS59892.2023.10330514