CSG664-CSG864 Reading/Research Courses

Prof. Patrick Wang, 260WVH, (617)373-3711, pwang@ccs.neu.edu


The main goal is to explore advanced Machine Learning with application to Pattern Recognition/Computer Vision/Artificial Intelligence/Web Design Topics in one of the following selected fields or closely related areas :

  • 3D Object Recognition and 2D Image Pattern Analysis

  • Thinning Methodologies (Parallel and Sequential)

  • Words (Characters) Learning/Understanding/Recognition Using Intelligent Techniques

  • Chinese OCR (Optical Character Recognition)

  • On-line 3D Imaging in Interactive Learning Virtual Environment using Internet and IM/Webcam Techniques

    • Reading references for this course will be mainly selected from the following publications and/or related papers/books :

  • ARTIFICIAL INTELLIGENCE: A Modern Approach , 2nd Edition, by Stuart Russel and Peter Norvig, Prentice Hall publisher. (2003), Chapters 18, 19, 21 in VI and 24 in VII
  • Artificial Intelligence, by Patrick Winston, Addison-Wesley (1997), Chapters 26, 27
  • Handbook of Pattern Recognition & Computer Vision, ed by C.Chen, L.Pau, P.Wang, WSP (1999), Chapters in Parts III & V
  • Handwritten Bank Check Recognition of Courtesy Amounts, International Journal of Imaging Graphics (IJIG), 2004 (with A.Gupta and R. Palacios) to appear [doc file]
  • Feedback based architecture for reading courtesy amounts on checks Int. J. Electronic Imaging (IJEI), (with A.Gupta, R.Palacios), 12(1) 194-202, 2003 [pdf file], [ps file]
  • 3D Vision and Object Recognition via Internet Web (with Yi, Min) [ppt], and [pdf] Int. J. Pattern Recognition and Artificial Intelligence (IJPRAI), 2001
  • High Level Visualization, Representation, Understanding, and Recognition of 3D Articulated Objects, by P. Wang The Encyclopedia of Microcomputers, ed. A. Kent, J. Williams, Marcer Dekker Pub. Co., 1999 [pdf file]
  • Analysis, Learning, and Recognition of Articulated Line Drawing Images, by P.Wang Int. J. CSIM, Special Issue on Image Recognition, 1998 v1-2 [pdf file]
  • Multimodal Interface for Human-Machine Communication, Advances of Hum an-Machine Communication, MPAI Series v.48, 9-11, 2002, WSP Pub (with P.Yuen and Y.Tang)

    • There will be no formal class schedules. Students will be expected to meet with the instructor for about one hour each week, to discuss selected papers, at mutually agreed upon time. If the student has any difficulties, the meeting time may be changed with advance notice. The final grade will be based on a term report, which will be due in the same week of the final exam. Note that the amount of work needed to be done for COM3805 (2 credit hours) is about half of COM3800 (4 credit hours)

    Week Contents
    I Introduction to the Course and Assign papers/articles
    II Student(s) read articles and discuss with instructor
    III Concentrate on a couple of articles to pursue
    IV Search for related articles/papers of the concentration
    V Draft a preliminary report summary
    VI Discuss why the focus, advantages, and what have learned so far
    VII Continue discussing the focus, advantages, and what have learned so far
    VIII Discuss assigned articles together with other related materials
    IX Continue discussing assigned articles, with other related materials
    X Comments on existing methodologies, their advantages vs disadvantages
    XI Continue comments on existing methodologies, their advantages vs disadvantages
    XII Discuss what might be improved, or any new ideas
    XIII Further discussions on what will be improved, or any new ideas
    XIV Discuss final conclusions, draft term report, and a short presentation
    XV (Final Exam Week) Term Report Due


    Back to Main Page
    Dr. Patrick S.P. Wang, Professor & IAPR Fellow
    College of Computer and Information Science, Northeastern University
    360 Huntington Avenue #260 WVH, Boston, MA 02115
    Phone: (617) 373-3711(Voice) --- Fax: (617) 373-5121
    Home Page: http://www.ccs.neu.edu/home/pwang
    pwang@ccs.neu.edu,
    For any comments, please send to one of the above email addresses. Thank you.