Instructor: Dr. Xiuwen Liu, email: liux@cs.fsu.edu, home page: http://www.cs.fsu.edu/~liux, office: 102C Carothers Building (MCH), phone: (850) 644-0050.

Course Home Page: http://www.cs.fsu.edu/~liux/courses/research/index.html
This web page contains the up-to-date information related to this class such as announcements, assignments, lecture notes, some useful links, and so on. I will have chapters and papers for reading assignments available from the web page.

Course Description: This course will cover the basic principles of computational modeling of visual perception, from color perception models, virtual contour models, texture models, and some computational models for recognition. Along with computational models for visual perception, this class will emphasize greatly on applications of those models, including content-based image retrieval, face recognition, and some techniques for object recognition. Students will have opportunities to learn and explore the state-of-art approaches in computer vision. Papers from the literature will be used as reading assignments. One or two programming projects based on student's own interests will be assigned.

Prerequisites: Some basic knowledge of algorithm designs and some experience in C/C++ programming. Email to the instructor if you have questions about the prerequisites.

Course Objectives: Upon successful completion of this course of study a student:

• knows basic problems in computational modeling of visual perception.
• understands color perception and its modeling, and knows its applications.
• understands the basics of texture modeling, and knows its applications.
• has some research experience in computer vision and image analysis.
• gains some knowledge of computer vision techniques in solving real-world problems.

Required Materials: Some chapters from Computer Vision -- A Modern Approach, Prentice Hall, 2000, by David Forsyth and Jean Ponce (Available at http://www.cs.berkeley.edu/~daf/book.html) and papers from the literature.

Tentative Schedule

• Week 1: Introduction.
• Week 2: Color perception and its modeling.
• Weeks 3 and 4:
  • Content-based image retrieval based on color histograms.
  • Color image classification and segmentation using color models
  • Tracking of moving objects based on color
• Week 5: Virtual phenonema such as virtual contours and its modeling.
• Week 6: Introduction to visual texture modeling.
• Week 7: Texture modeling based on Markov Random Field Models.
• Weeks 8 and 9: Spectral histogram as a generic texture feature and Julesz ensembles for textures.
• Weeks 10 and 11: Overview of texture classification techniques and classification based on spectral histograms.
• Week 12: Content-based image retrieval based on texture characteristics.
• Week 13: Introduction to image sequence analysis, including optical flow estimation, motion segmentation, and 3-D reconstruction.
• Week 14: Introduction to object recognition and face recognition in particular.
• Week 15: Directions for future research.

Contact information: Prospective students are encouraged to visit Dr. Xiuwen Liu's web page at http://www.cs.fsu.edu/~liux more information on his current research interests and projects. Please contact Dr. Xiuwen Liu at liux@cs.fsu.edu or (850) 644-0050 for questions, comments, or further information.

Interesting examples: The following are a few examples of what will be covered in the class.

Virtual contours: These visual phenomena reveal some of the interesting characteristics of visual perception.

Visual textures: How to characterize different kinds of textures.

Face recognition: How to characterize different faces and recognize them.

Motion analysis: How to estimate the motion of the player and synthesize sequences from different viewpoints.