E1 216 : Computer Vision
LECTURE 05: ROBUSTNESS
2024
In this lecture we shall consider robustness issues in computer vision. In particular we shall consider the RANSAC method and M-estimation using IRLS.
Notes:
- Required Reading: Section 8.1.4 of Szeliski's book (Second Edition)
- Required Reading : Section 10.4 of Forsyth and Ponce (Second Edition)
- Suggested Reading: "Robust Techniques for Computer Vision" There are many recent papers on using deep learning in various ways to achieve robustness.