Presentation

DescriptionThis course will survey the state-of-art in computational displays, computational imaging and perceptual graphics geared towards virtual and augmented reality devices. In addition, a live coding session will introduce various optimization methods for optimizing displays, cameras and perception. The course intends to capture people from beginner to advanced stages.

Prerequisites Experts in Virtual Reality and Augmented Reality. We believe virtual and augmented reality experts will enormously benefit from the survey and overview provided by our keynotes and domain experts. Newcomers. We trust that our interactive sessions following the survey talks can help a newcomer (e.g., graduate student, interested people) to grasp the concepts fast. In addition, we trust that the newcomer attendees will be able to optimize their first hologram for a standard holographic display, reconstruct an image from a lensless camera capture and use perceptual guidance in their future optimizations. General SIGGRAPH audience. The topics introduced in this course are highly relevant to SIGGRAPH's general audience as it describes critical hardware related ingredients for building the future's immersive three-dimensional environments. Specifically, our course is at the core of images and visuals, a primary research cause in computer graphics.

Intended Audience The material for this course assumes some basic assumptions about the course participants. First, the participants in the audience should have some familiarity with concepts such as Metaverse, Telelife, virtual reality and augmented reality. Participants can be from various technical backgrounds but are willing to advance their understanding of the optimization side of displays, cameras or perception. The course material will describe the base hardware used in the demonstrated optimizations. However, the lecturers will not cover how to build this hardware from the ground up. Instead, they will provide a quick overview and cite relevant resources from the most recent literature. The attendees are also expected to be knowledgeable or willing to learn Python programming language, modern machine learning libraries, signal theory, and optics. But most importantly, above-average interest in optimizing future’s devices is a must. The code material of this course is a result of various research works from the Computational Light Laboratory that was conducted in the passing one-year timeframe (2021-2022). These research works include new methods for foveated rendering and image statistics, learned techniques for holographic light transport, learned optimizations for multiplane holography, perceptually guided hologram generation routines in displays and learned optimizations for lensless cameras. Individuals willing to replicate the outcome of our materials from this course on their local machines have to install several pieces of software in their operating systems. Such individuals must be familiar with the Python programming language, and they should install Python and Jupyter Notebooks in their operating systems. In addition, these individuals will need to install Torch with its Python bindings, Matplotlib and plotly libraries for plotting purposes while using the provided Jupyter Notebooks. When we compiled this material, our production machines used the Python distribution 3.9.7, Torch distribution 1.9.0, Matplotlib distribution 3.3.4 and Jupyter Notebook distribution 6.2.0. As a final piece, please make sure to install our library using: pip install odak The participants willing to go beyond this course and learn more about relevant research are welcome to reach out to the lecturers via email. In addition, Computational Light Laboratory offers seminars from experts on relevant topics. Finally, computational Light Laboratory also invites all attendees to a research hub formulated as a Slack Group. This way, curious readers and attendees of our course can keep up-to-date and meet more folks in the relevant fields.

Lecturers

Koray Kavaklı

Koç University

University College London (UCL)

David Robert Walton

University College London (UCL)

Nick Antipa

University of California San Diego

Rafał Mantiuk

University of Cambridge

Douglas Lanman

Reality Labs Research, Meta

University of Washington

Kaan Akşit

University College London (UCL)