Ron Fedkiw
(Full) Professor
Stanford Computer Science

Ph.D. Applied Mathematics, UCLA

LEFT PHOTO: circa 2005; RIGHT PHOTO: circa 2017

Computer Science Department
Stanford University
Gates Computer Science Bldg., Room 207
Stanford, CA 94305-9020

Special Effects, Machine Learning, and You...
I'm not really a graphics person so much as I'm a Hollywood special effects person. These days my work on special effects focuses quite a bit on face and body animation and simulation, trying to outwit the uncanny valley. Traditionally, one used only computer vision techniques for this sort of work, but we're now successfully mixing in quite a bit of physical simulation as well. In fact, I have many ongoing projects mixing in real world data and simulation in order to create more realistic simulations for Hollywood special effects (cloth is next).

It turns out that students interested in Hollywood special effects with backgrounds in math, physics, computer vision, and machine learning (all combined) are hard to find...

CS205A and Machine Learning

  • Almost 20 years ago I took over CS205 and revamped it from a robotics/vision course to be more of a continuous math course and to include applications important to computer graphics.
    The time has come to revamp CS205A to consider the plethora of traditional applied math and continuous mathematics that lie at the heart of machine learning and deep learning.
    This is especially important as machine learning and deep learning infiltrate every area of computer science, as well as biology, finance, etc.

  • At some point along the way, we changed the name of the course to include graphics, changed the numbering from CS205 to CS205A, created new homework and projects, etc.
    All this will happen again starting Winter quarter 2019 (a bit more than 1 year from now) when we will relaunch CS205A as CS205L: Continuous Mathematics with Applications to Machine and Deep Learning.
    The course will still cover linear algebra, optimization, etc., but will focus on these mathematics in the special ways in which they are important for machine learning and deep learning.
    For example, instead of spending a lot of time covering ODEs for computational mechanics applications, we will cover them from the standpoint of momentum methods for training neural networks with steepest decent.
    Also, instead of finishing the course with some brief ideas on PDEs for fluids in graphics, we will finish with a discussion of network design from a mathematical viewpoint, including CNNs, RNNs, manifolds, etc.

  • I am looking for any interested students to work with Winnie Lin, my head CA, and myself on anything related to this relaunch of CS205.
    Anticipating potentially large enrollments, we would like to be amply prepared before the course launches with homework, exams, a syllabus, ideas for content and lectures, slides, etc.
    If you are interested in any way, please email Winnie ( so that she may put your name on a mailing list.

  • You are also welcome to email Winnie and get on the list even if you are just interested in taking the course and want to hear about its development.
    CS205 has always been a service class, so hearing from those interested in taking such a class is super-important to us.

  • CS205L: Continuous Mathematical Methods with an Emphasis on Machine Learning

    A survey of numerical approaches to the continuous mathematics used in computer vision and robotics with emphasis on machine and deep learning. Although motivated from the standpoint of machine learning, the course will focus on the underlying mathematical methods including computational linear algebra and optimization, as well as special topics such as automatic differentiation via backward propagation, momentum methods from ordinary differential equations, CNNs, RNNs, etc. (Replaces CS205A, and satisfies all similar requirements.) Prerequisites: Math 51; Math 104 or 113 or equivalent or comfort with the associated material.

    Brief Bio
    Fedkiw received his Ph.D. in Mathematics from UCLA and spent part of his postdoctoral studies at Caltech in Aeronautics before joining the Stanford Computer Science Department. He was awarded an Academy Award from the Academy of Motion Picture Arts and Sciences (twice: 2008 and 2015), the National Academy of Science Award for Initiatives in Research, a Packard Foundation Fellowship, a Presidential Early Career Award for Scientists and Engineers (PECASE), a Sloan Research Fellowship, the ACM Siggraph Significant New Researcher Award, an Office of Naval Research Young Investigator Program Award (ONR YIP), the Okawa Foundation Research Grant, the Robert Bosch Faculty Scholarship, the Robert N. Noyce Family Faculty Scholarship, two distinguished teaching awards, etc. He has published over 120 research papers in computational physics, graphics, and vision, a book on level set methods, and is currently working at the interface between physical simulation and machine learning - recently joining the Stanford Artificial Intelligence Laboratory (SAIL). Currently, he serves on the editorial board of the Journal of Computational Physics. For the past 18 years, he has been a consultant with Industrial Light + Magic, receiving screen credits on movies such as "Terminator 3: Rise of the Machines", "Star Wars: Episode III - Revenge of the Sith", "Poseidon", "Evan Almighty", "Kong: Skull Island", etc. Of all his achievements, he is most proud of the combined accomplishments of the 30 Ph.D. students that he has graduated so far, and looks forward to seeing what the next 30 will do!



    Computational Physics...

    Computer Graphics, Vision & Biomechanics...


    Ph.D. Students

    Former Ph.D. Students Former Postdoctoral Scholars

    A Note on Rejected Papers

    All too often young researchers get discouraged when they receive peer reviews that are incorrect, misinformed, or all too often merely intended to silence the authors and their ideas. Personally, I have always been amazed that academics who devote their lives to producing new information actually work to censure and diminish the work produced by others, and often take pride in doing just that. As time goes on, one learns to distinguish between those in academia who love the work and those that have instead turned academia into some sort of career aggressively optimizing their stature at the expense of the community as a whole. For young researchers this can be quite daunting, but I strongly encourage you to stick to your ideas and goals and the pursuit of what interests you. Remember, the content of your paper and the value of its ideas are not diminished because it was rejected from your preferred venue. The content of the paper itself does not change because of the name of the journal printed on the upper corner of the page! To emphasize this, I decided to list my 3 most cited REJECTED papers along with their google scholar citation counts:
  • "A Boundary Condition Capturing Method for Multiphase Incompressible Flow", 618 citations, rejected from J. Comp. Phys.
  • "Simulation of Clothing with Folds and Wrinkles", 515 citations, rejected from Siggraph
  • "Fast Surface Reconstruction using the Level Set Method", 469 citations, rejected from Siggraph