Instructor: Dr.
Alper Yilmaz
(yilmaz at cs.ucf.edu)
Room:
CSB 250
Office Hours:
3.30pm-4.30pm Monday
Wednesday
Grader:
Imran Junejo
(ijunejo at cs.ucf.edu)
Course Syllabus |
- Introduction and course objectives (January 9,
2006)
- Lecture
2 (January 11, 2006)
- Shape representations
- Appearance representations
- Medial axis transform
- Programming
assignment: Implementation
of the medial axis transform
- Due
date: January
25, 2006
- Sample
image1, Sample
image2, Sample
image3
- Deliverables:
- Report: The report
should include input image, several intermediate iterations (scaled up
to visuallize), Final medial axis in the form of a skeleton.
- Hardcopy of source code.
- January 16 - Martin Luther King Day
- Lecture 3
(January 18, 2006)
- Visual image features
- Color, texture, edge, optical flow
- Edge detection
- Lukas Kanade optical flow
- Texture feature extraction
- Lecture 4
(January 23, 2006)
- Object detection
- Point detectors
- Mathematics of Harris
- KLT detector
- Everyone is required to attend the Distinguished
Vision Lecturer Series
- Homework: Algebraically show the rotation
invariance of the Harris corner detector.
- Due
date: February 1, 2006
- Lecture 5
(January 25, 2006)
- Object detection
- Point detectors
- SUSAN detector
- SIFT Key Point Detector
- 1st
Programming Project: Implementation of SIFT Key point
detector
- Due
date: February
27, 2006
- Lecture 6
(January 30, 2006)
- Object detection
- Point detectors
- SIFT Key Point Detector cnt'd.
- Background Subtraction
- Frame differencing
- Single Gaussian
- Mixture of Gaussians
- Programming
assignment: Implementation
of background subtraction using single gaussian per pixel
- Due
date: February
13, 2006
- Sample
Sequence You
need to train background for first 120 frame, then
- Deliverables:
Report: The
report should include several images from input sequence, mean
background model as an image, standard deviation of background model
as image, object coarse detection results (no additional
operations such as morphological operators).
Hardcopy of source code.
- Lecture 7
(February 1, 2006)
- Object detection
- Global motion Compensation
- Homework
assignment: Derive bilinear and pseudo
perspective transformations (see slides)
- Due
date: February
8, 2006
- Lecture 8
(February 6, 2006)
- Image Segmentation
- Human perception
- Clustering and K-means
- Lecture 9
(February 8, 2006)
- Machine learning
- Support Vector Machines
- Presented by Vladimir Reilly
- Homework
assignment: Derive
bilinear and pseudo perspective transformations using Taylor series
expansion (if you have done so in the previous assignment you do not
need to do it again)
- Due
date: February
15, 2006
- Lecture 10
(February 13, 2006)
- Machine learning
- Support Vector Machines (continued)
- Adaptive Boosting
- Presented by Alex Kachurin
- Lecture 11 (February 15, 2006)
- Machine learning
- Adaptive Boosting (continued)
- Lecture 12
(February 20, 2006)
- Image Segmentation
- Minimum Cut and Normalized Cut
- Presented by Paul Scovanner
- Sample Results: 1
2
- Lecture 13 (February 22, 2006)
- Class discussion on SIFT implementation
- Lecture 14 (February 27, 2006)
- Image Segmentation
- Convolving two Gaussians
- Reading:
- Fukunaga, K., Hostetler, L.: The estimation
of the gradient of a density function,
with applications in pattern recognition. IEEE Trans. Information
Theory 21 (1975)
32–40.
- Comaniciu, D., Meer, P.: Mean shift: a
robust approach toward feature space anal-
ysis. IEEE Trans. on PAMI (2002) 603-619.
- Homework
assignment: 1)
Show that the derivative of the Kernel used in KDE is as given in slide
18 of this lecture. 2) Show the equality at the end of derivation here.
- Due
date: March 20, 2006 (UPDATED)
- Lecture 15 (March 1, 2006)
- Image Segmentation
- Mean-Shift Clustering (continued)
- Project Program Demos
- Lecture 16 (March 20, 2006)
- Snakes, active contours: formulation
- Reading
- M. Kass, A. Witkin, and D. Terzopoulos. Snakes:
Active contour models. In Proc. 1st ICCV, pages 259-268, June 1987.
London, UK.
- Yongjik Kim. A summary of Implicit Snake Formulation.
- Trucco & Verri, "Introductory Techniques for 3D Vision", Prentice Hall (Chapter 5.4)
- Programming
assignment: Implementation
of object tracking using SIFT operator
- Due
date: April 3, 2006
- Deliverables:
Report: The
report should include trajectories of the tracked key points.
Hardcopy of source code.
- .Lecture 17 (March 22, 2006)
- Active contours, level set formulation
- Lecture 18 (March 27, 2006)
- Level set formulation (continued)
- 2nd Programming Project: Implementation of image segmentation using Level Set based contour evolution
- Due
date: April
24, 2006 (Last day of classes)
- N. Paragios and R. Deriche. Geodesic active regions and level set methods for
supervised texture segmentation. The International Journal of Computer
Vision, 46(3):223-247, 2002
- Deliverables:
- Source code on a CD,
- Project report discussing problems
encountered, where the algorithm fails,
- Images of intermediate contour evolution
- Classroom demo is required on an unknown
set of images
- Program should display evolving contour
- Lecture 19 (March 29, 2006)
- Active Contours
- Programming
assignment: Implementation
of texture modeling using histograms of Gabor wavelet's spectrum analyzer. You need to use 4 Gabor wavelets for this purpose.
- Due
date: April
10, 2006
- Deliverables:
Report: The
report should include several texture images, and the resulting historams along with weights.
Hardcopy of source code.
- Lecture 20 (April 5, 2006)
- .Lecture 21 (April 10, 2006)
- Veenman, C., Reinders, M., and Backer, E. 2001.
Resolving motion correspondence for densely moving points. PAMI 23, 1,
54–72.
- Presented by Pavel Babenko
- Programming
assignment: Implementation
of Level Set Function. You need to generate the level set surface from an image containing the object mask
- Due
date: April
17, 2006
- Deliverables:
Report: The
report should include several mask images, and the resulting level set function.
Hardcopy of source code.
- Lecture 22 (April 12, 2006)
- .Lecture 23 (April 17, 2006)
- Project Demo (April 24, 2006)
April 24 Classes End
- Adeel Ali: R. Vidal, Y. Ma and S. Sastry, 2003, Generalized Principal Component Analysis (GPCA). CVPR, 621--628
|
Assignments,
projects etc.
- 1st
Programming assignment: Implementation of the medial
axis transform. Due date: January 25, 2006
- 1st
Homework: Algebraically
show the rotation invariance of the Harris corner detector. Due date: February
1, 2006
- 1st
Programming Project: Implementation of SIFT Key point
detector. Due date: February 27, 2006
- 2nd
Programming assignment: Implementation of
background subtraction using single gaussian per
pixel. Due date: February 13, 2006
- 2nd
Homework: Derive
bilinear and pseudo perspective transformations (see slides lecture 7). Due date: February
8, 2006
- 3rd
Homework: Derive
bilinear and pseudo perspective transformations using Taylor series
expension (see slides lecture 7) (if you have done so in the
previous assignment you do not need to do it again). Due date: February
15, 2006
- 4th
Homework: 1) Show that the derivative of the Kernel used in KDE is as given in slide 18 of lecture 14. 2) Show the equality at the end of derivation here. Due date: March 20, 2006
- 3rd Programming assignment: Implementation of key point tracking using SIFT. Due date: April 3, 2006
- 2nd Programming Project: Implementation of level set based segmentation. Due date: April 24, 2006
- 4th Programming assignment: Implementation of texture modeling. Due date: April 10, 2006
- 5th Programming assignment: Implementation of generation of level set function from a mask image. Due date: April 17, 2006
|
Matlab Tutorials:
|
|
|