Course 67684

Advanced Topics in Photorealistic Computer Graphics

room image

Instructor: Dr. Dani Lischinski
Time & Place: Tuesday, 12:00-13:45 (Shprinzak 201)
Credit: 3 credits

The first part of the course will concentrate on global illumination algorithms capable of producing photo-realistic images, such as the one shown above. We shall cover (in depth) finite-element based methods, such as radiosity and its derivatives, as well as ray tracing and Monte Carlo based techniques for physically-based simulation of illumination in 3D scenes.

The second part of the course will concentrate on image-based modeling and rendering. We shall cover some of the image warping and image morphing algorithms that have recently become popular due to their use for special effects in motion pictures, as well as other algorithms that synthesize new images from existing ones.

Finally, we'll discuss some of the recent state-of-the-art research results in photorealistic rendering.

Requirements: Those who take the course for credit will be required to complete three programming assignments. In these assignments students will implement some of the algorithms covered in the course. Programming will be done is small teams (2-3 students). Some of the code will be provided to students in the form of C++ libraries, so that students will be able to concentrate on the algorithmic aspects of the assignments. Hands-on experience (via programming and debugging) is the best way of gaining true understanding of the material in this course. However, students who wish only to attend the lectures (without credit) are also welcome.

Prerequisites: the course is open to anyone who has taken the introductory course in computer graphics (67609), or have had equivalent exposure to computer graphics. In particular, we will assume familiarity with the basics of the 3D rendering pipeline, basics of color science, and simple shading models.


Online Material:

  1. Introduction (slides: Intro.pdf)
  2. Rendering concepts: light, radiance, reflectance (slides: Concepts.pdf, lecture notes: notes2.pdf)
  3. Derivation of the Radiosity method (lecture notes: notes3.pdf)
  4. Hierarchical Radiosity (lecture notes: hr.pdf)
  5. Monte Carlo methods (slides: MonteCarlo.pdf)
  6. Optimally combining Monte Carlo estimators (slides: OptimalCombining.pdf)
  7. Photon maps (slides: PhotonMaps.pdf)
  8. Image-Based Rendering (slides: IBR-slides.pdf)
  9. Light Field Rendering (slides: LightFields.pdf)
  10. Image-Based Relighting (slides: Relighting.pdf)

Programming Assignments:

  1. Hierarchical Radiosity: ex1.txt (Date due: May 4, 2004)
  2. Monte Carlo Integration: ex2.txt (Date due: June 8, 2004)
    See example results here
  3. LDI warping: ex3.txt (Date due: July 15, 2004)