CSC2522S: Advanced Image Synthesis


Instructor: Eugene Fiume.
Lectures: 11-1pm, Thursdays -- BA5187 (dgp lab)

This course is an in-depth study of the mechanisms in computer graphics that allow us to make beautiful, realistic images. The main topics we will be covering are realistic illumination models (whether physically-based or otherwise), global illumination and light transport, and ray-tracing and path-tracing algorithms, participating media, texture and environment mapping, sampling and filtering.

The grade will be based on a seminar and a project.

Some References

  1. We will be referring to the book Physically Based Rendering by Matt Pharr and Greg Humphreys.
  2. The home page for the pbrt software and scenes.
  3. My notes:
    1. An introduction to radiometric terms, light transport.
    2. An primer on sampling and filtering, the sampling theorem, reconstruction.
  4. The Stanford image synthesis slides notes, orginally by Pat Hanrahan but refined by many for the Stanford image synthesis course are quite informative. These are excellent course notes with material analogous to our coverage.
  5. A good introduction to the Monte Carlo techniques we are discussing is by Eric Veach and others. There are many other interesting notes there.
  6. A more intensive but very readable coverage of physically based image synthesis with an emphasis on path tracing and Monte Carlo methods (including Metropolis Light Transport) is Eric Veach's thesis.
  7. The SIGGRAPH and Eurographics Symposium on Rendering Proceedings.


Student Talks

  1. TBA