Is Simulation (still) the Future of Rendering, or will it be Capture and Learning?

Speaker:  George Drettakis – Sophia-Antipolis, France
Topic(s):  Graphics and Computer-Aided Design


In this talk we will examine recent evolutions in traditional, simulation-based rendering, and those in capture and learning based solutions. Based on several concrete examples, we will discuss the fundamental issues behind the contemporary quest for visual realism, which, as always, relate to accurate and realistic visual rendition of geometry, materials and illumination. We will first discuss the enormous progress of traditional forward simulation based methods, especially for illumination algorithms. We will then consider one remaining major challenge, which is the provision of accurate geometry and materials needed for these algorithms to run, and the fact that this is currently still done with manual labor and heuristics. We will see how recent work using capture and learning try to overcome this challenge, and discuss whether in some cases such approaches actually render illumination algorithms superfluous. We will close with questions on how these seemingly opposing approaches -- simulation vs. capture and learning -- could potentially come together, bringing us closer to the holy grail of easy-to-create, completely realistic imagery in real-time.

About this Lecture

Number of Slides:  50
Duration:  120 minutes
Languages Available:  English, French, Greek
Last Updated: 

Request this Lecture

To request this particular lecture, please complete this online form.

Request a Tour

To request a tour with this speaker, please complete this online form.

All requests will be sent to ACM headquarters for review.