Syllabus
This course is based on notes and articles from the literature. Additional materials will be posted below as appropriate. Text in bold on this page is typically a link.
This syllabus is a plan, not a commitment. Depending on class interest and the time needed to cover the various topics, it may be necessary to skip some of the topics below. The topic in color is the topic currently being covered. Topics and materials below the topic in color may change. Materials in parentheses (not square brackets!) are optional.
Paper references in the topics below are specified through their Digital Object Identifier, when available, or through a link that recognizes Duke affiliation. These links will get you to the full article if you or your institution have proper access privileges. For Duke students, this typically means that the link will work from a Duke computer, but not from elsewhere.
| 0 | Overview and Logistics |
| Date | Topic | Supplementary Materials |
|---|---|---|
| Aug 30 | Contents and logistics of the course | Fei-Fei Li's TED talk. Intro to LaTeX. Linear Transformations. |
| 1 | Image Processing |
| Date | Topic | Supplementary Materials |
|---|---|---|
| Sep 1 | Convolution and Filtering | — |
| Sep 6 | Image Differentiation | — |
| Sep 8 | Image Pyramids | Matlab image processing code |
| Sep 13 | Math Corner: The Singular Value Decomposition | — |
| 2 | Object Detection |
| Date | Topic | Supplementary Materials |
|---|---|---|
| Sep 15 | Histograms of Oriented Gradients | Dalal & Triggs |
| Sep 20 | Supervised Learning | Domingos |
| Sep 22 | [Supervised Learning, cont'd] | — |
| Sep 27 | Random Forest Classifiers | (Criminisi et al.) |
| Sep 29 | Pedestrian Detection | Hough Forests |
| 3 | Convolutional Neural Nets |
| Date | Topic | Supplementary Materials |
|---|---|---|
| Oct 4 | Math Corner: Function Optimization | — |
| Oct 6 | Convolutional Neural Nets | LeCun, Bengio, and Hinton |
| Oct 13 | [Convolutional Neural Nets, cont'd] | (Vedaldi and Zisserman's CNN Practical) |
| Oct 18 | Training Convolutional Neural Nets | — |
| Oct 20 | Convolutional Neural Nets for Image Recognition | Krizhevsky et al. |
| 4 | Point Features |
| Date | Topic | Supplementary Materials |
|---|---|---|
| Oct 25 | [In-Class Midterm Exam] | — |
| Oct 27 | Math Corner: Linear Systems | — |
| Nov 1 | Point Correspondences (sections 1-3 of the notes) | — |
| Nov 3 | Points of Interest (section 4 of the notes) | — |
| 5 | Image Formation |
| Date | Topic | Supplementary Materials |
|---|---|---|
| Nov 8 | Rigid Geometric Transformations | — |
| Nov 10 | A Camera Model | — |
| Nov 15 | Epipolar Geometry (section 1 of the notes) | — |
| Nov 17 | The Essential Matrix (section 2 of the notes) | — |
| 6 | 3D Reconstruction |
| Date | Topic | Supplementary Materials |
|---|---|---|
| Nov 22 | The Eight-Point Algorithm | (Longuet-Higgins) |
| Nov 29 | Camera Calibration | (Zhang; rodrigues.m, initialize.m) |
| Dec 1 | The Standard Reconstruction Pipeline | (Rome in a Day, Multi-View Stereo) |
COMPSCI 527, Duke University, Site based on the fluid 960 grid system