|
CS484: Introduction to Computer Vision
Fall 2018
|
Instructor |
Min H. Kim, [Room] 3429, [email] |
Course description
|
This course provides a comprehensive introduction to low-level computer vision, including the foundations of camera image formation, geometric optics, feature detection, stereo matching, motion estimation, image recognition, scene understanding, etc. This course will help students develop intuitions and mathematics of various computer vision applications. |
Lecture time and place |
Monday and Wednesday 1:00PM—2:30PM, Rm. 1501, Bldg. E3-1 |
TA office hours |
Monday and Wednesday 6:00PM—8:00PM, Rm. 402, Bldg. N-1
Thursday 2:30PM—5:30PM, Rm. 404, Bldg. N-1
|
Teaching assistants |
Suk Jun Jeon, [e-mail]
Hyunho Ha, [e-mail]
Hyeonjoong Jang, [e-mail]
Shinyoung Yi, [e-mail]
Inseung Hwang, [e-mail]
|
Reference books |
Reinhard Klette (2014) Concise Computer Vision, Springer [site]
Richard Szeliski (2010) Computer Vision: Algorithms and Applications, Springer [site] |
Prerequisites |
There are no official course prerequisites. |
Course goal |
Student will establish theoretical and practical foundations of computer vision and be familiar with various computer vision applications. |
Tentative schedule |
(Note that this curriculum will be revised adaptively.) |
| |
Week |
Date |
Lecture |
Readings |
Slides |
HW |
Due |
| |
1 |
08/27 |
Introduction to computer vision |
Klette: 1,
Szeliski: 2.2, 2.3 |
slide01 |
|
|
| |
|
08/29 |
Light and color |
Klette: 1,
Szeliski: 2.2, 2.3 |
slide02 |
|
|
| |
2 |
09/03 |
Color vector and transformation |
Klette: 1,
Szeliski: 2.2, 2.3 |
slide03 |
|
|
| |
|
09/05 |
Color space and gamma correction |
Klette: 1,
Szeliski: 2.2, 2.3 |
slide04 |
|
|
| |
3 |
09/12 |
Convolution and cross-correlation |
Klette: 1, 2, Szeliski: 3.2 |
slide05 |
HW1 |
9/16,23:55 |
| |
4 |
09/17 |
Spatial vs. frequency domain |
Klette: 1, 2, Szeliski: 3.4 |
slide06 |
HW2 |
9/30, 23:55 |
| |
|
09/19 |
Fourier series and Fourier transforms |
Klette: 1, 2, Szeliski: 3.5.2, 8.1.1 |
slide07 |
|
|
| |
5 |
09/24 |
Mid-Authumn Festival Days |
|
|
|
|
| |
6 |
10/01 |
Fourier transforms and convolution Theorem |
Klette: 1, 2, Szeliski: 3.5.2, 8.1.1 |
|
|
|
| |
|
10/03 |
Edge detectors and image pyramids |
|
|
HW3 |
TBA |
| |
7 |
10/8 |
Camera: geometric optics of a digital camera, image formation model |
|
|
|
|
| |
|
10/10 |
Image analysis: line and circle detection and 2D shape analysis |
|
|
HW4 |
TBA |
| |
8 |
10/15 |
Mid-term exam (TBA) |
|
|
|
|
| |
9 |
10/22 |
Motion analysis: 3D motion and 2D optical flow |
|
|
|
|
| |
|
10/24 |
Image segmentation: image and video segmentation |
|
|
|
|
| |
10 |
10/29 |
Computational photography: light-field imaging |
|
|
|
|
| |
|
10/31 |
Computational photography: HDR imaging |
|
|
HW5 |
TBA |
| |
11 |
11/05 |
3D shape reconstruction: surface topology and structured lighting |
|
|
|
|
| |
|
11/07 |
3D shape reconstruction: photometric stereo, Epipolar geometry, RANSAC and SfM |
|
|
HW6 |
TBA |
| |
12 |
11/12 |
Stereo matching: stereo vision, matching cost, dynamic programming matching |
|
|
|
|
| |
|
11/14 |
No lecture (Microsoft Faculty Summit) |
|
|
|
|
| |
13 |
11/19 |
Object detection: Haar Wavelets and random decision forests |
|
|
|
|
| |
|
11/21 |
Recognition overview and bag of features |
|
|
|
|
| |
14 |
11/26 |
Large-scale instance recognition |
|
|
|
|
| |
|
11/28 |
Machine learning: supervised learning vs. unsupervised learining |
|
|
|
|
| |
15 |
12/03 |
Convolutional neural networks (CNN) |
|
|
|
|
| |
|
12/05 |
No lecture (SIGGRAPH Asia 2018) |
|
|
|
|
| |
16 |
12/10 |
Final exam (TBA) |
|
|
|
|
| |
|
|
|
|
|
|
|
|
|
|
|
Hosted by Visual Computing Laboratory, School of Computing, KAIST.
|
 |
|
