CSCI-737: Pattern Recognition (Spring 2016)

  RIT Department of Computer Science

CSCI-737-01 Pattern Recognition (Spring 2016)

Instructor: Prof. Richard Zanibbi (web page; rxzvcs at rit dot edu )
Office Hrs: M 1-2pm, F 1-2pm, 3-4pm
Lectures: MWF 2-2:50pm, GOL-3560

[ News ] --- [ Schedule ] --- [ Syllabus ] --- [ Resources ] --- [ MyCourses ]


News


Week 16

  • Next Week
    • Office Hours Next Week: MW 1-3pm. Send email if you wish to meet but have a conflict with these times.
    • Project Team Presentation Slides are on MyCourses under "Projects"
    • Benchmark/Bonus results will be posted on MyCourses as they are ready.
  • This week students will present their project summary talks (Assignment 5). The schedule for presentations is listed below.
  • Project 3 is due Sunday May 22nd at 11:59pm. Aim to finish Thursday May 19th if you can!
  • Please complete the online course evaluation. I would particularly appreciate comments on what worked in the course, and what might be improved in the future.

Week 15

  • 3rd Extension for Project 2. Project 2 is due Wednesday at 11:59pm. There will be *no* extension after this point.
  • The final round of paper presentations will be given on Friday, on parsing handwritten math.
  • The final assignment (the project presentations) will be posted on Wednesday.
  • The final part of the Project will be posted on Wednesday, and will be due Sunday May 22nd at 11:59pm (*hard* deadline due to grading requirements). Aim to finish the project by Thursday May 19th (before commencement and convocation).
  • Schedule for Presentations Next Week:
    • Monday (3): Chris and Juraj, Elizabeth and Evan, Deepali and Michael, (+ one research paper presentation)
    • Wednesday (5): Anurag and Swanand, Mandy and Michael, Lakshmi and Himesh, Ying-Kai and Brendan, Winston and Jefferson
    • Friday (4): Tobin and Yu, Kardo and Victor, Prasanth and Srinath, Suhas

Week 14

  • 2nd Extension for Project 2: Proj. 2 is due Sunday at 11:59pm.
  • Papers for the last set of paper presentation (Parsing, part II, by students in the second group of presenters) have been posted. Doodle poll: paper selection.
  • In Week 16 (two weeks from now), on MWF each project group will give a 10 minute presentation of their system, including a sketch of how they plan to parse expressions. The project group presentations will be the fifth assignment in the course (i.e. worth 5% of your final grade).

Week 13

  • Assignment 4 is due Friday at 11:59pm.
  • Project 2 is due Wed. Apr. 27th at 11:59pm.
  • Parsing Paper Presentations. Students in the first group of paper presenters will give their final paper presentations on Friday (Week 13).
  • CROHME 2016: For those interested, the next CROHME competition has started (results are due in mid-June). Web page: CROHME 2016.

Week 12

  • Parsing Paper Presentations. Students in the first group of paper presenters (who presented last Friday) will give their final paper presentations next Friday (Week 13). This will be the first set of papers on parsing. Doodle poll to select papers is located here.
  • Deadline Extension: Project 2 is now due Wednesday, April 27th at 11:59pm. Make sure to use the updated version of the LgEval library available through MyCourses.
  • Assignment 4 is due Friday Apr. 22nd at 11:59pm.
  • Slides from presentations in previous weeks have been posted on MyCourses under 'Research Papers.' You can find the slides under the corresponding list of papers for presentation.

Week 11

  • The second set of paper presentations will be given in-class on Friday.
  • Project 2 has been posted on MyCourses. The project is due Sunday, April 24th at 11:59pm. Please make sure to use the updated LgEval library.
  • Project 1 Extension: Project 1 can be submitted until Tuesday at 11:59pm without penalty.

Week 10

  • Project 1 is due Sunday, April 1st.
  • Papers for presentation next Friday have been posted on MyCourses. The doodle poll to select papers for presentation is located here: Doodle poll. Students from the 1st group (groups 1-4 below) will present.
  • Project 2 will be posted early next week.
  • The grading key and notes for Assignment 3 are available under the "Assignments" on MyCourses.
  • Of interest: here is a web page describing some early handwriting recognition and pen-based computing systems, along with some interactive demonstrations of common pre-processing steps for stroke data: Back to the Future of Handwriting Recognition.

Week 8

  • Assignment 3 is due on Wednesday at 11:59pm. You will want to look at the project description as well as the assignment write-up.
  • We will have our first round of paper presentations on segmentation this Friday.
  • Project 1 is due Sunday, April 1st.
  • Illustration of eigenvectors and eigenvalues: Wikipedia demonstration.

Week 7

  • Assignment 3 has been posted, and is due next Wednesday at 11:59pm.
  • The first set of papers on segmentation have been posted on Doodle: paper selection. Papers will again be assigned on a first-come, first-serve basis. The last four groups that presented last Friday will present next Friday (Week 8).
  • Project 1 has been posted on MyCourses, and is due Sunday, April 3rd. Teams of two students should sign up using this doodle poll: project teams. The papers themselves are available through MyCourses.

Week 6

  • Class on Wednesday: on Wednesday (Career Fair Day), we will go over Assignments 1 and 2, and have a first look at tools for the course project (time permitting).
  • Paper presentations: we will have our second round of paper presentations on Friday, on recognizing handwritten math symbols.

Week 5

  • Assignment 2 Extension the assignment is now due Wednesday at 11:59pm.
  • Project 1 (classifying handwritten symbols) will be posted next week. Projects will be completed by teams of two students, and will be due before Spring Break.
  • Papers for the second set of presentations will be posted on Friday.
  • Read the two Quinlan C4.5 documents (Chs. 1-2, and 4, on decision tree learning, splitting features, gain ratio, and pruning using pessimistic error estimation)
  • Read the Criminisi paper on Random Forests, Sections 2.3.5 - 3.3.

Week 4

  • Assignment 2 has been posted on MyCourses. It is due next Tuesday at 11:59pm.
  • The first set of research paper presentations will be given in-class on Friday. Instructions are provided below. Papers for the second set of classification papers will be posted next Friday (Wk 5).
  • Read Chs. 1 and 2 of the Quinlan C4.5 book (available through MyCourses).

Week 3

  • Assignment 1 is due Monday at 11:59pm. Submit your write-up and program through the MyCourses drop box.
  • Papers for the first set of research paper presentations have been posted on MyCourses under "Content -> Research Papers." Students in the first 4 groups listed in the image below will present a paper next Friday (Week 4).
    • Papers will be assigned on a first-come, first serve basis. The online poll to select a paper is located here.
    • Slides: One student from each group should submit the presentation slides as a .pdf file through the dropbox in MyCourses (labeled "Classification Slides"). The title slide should provide the full paper citation, with the paper title, author names, bilbiographic information (journal, volume, issue number, pages, etc.), and the names of your group members.
    • Students are expected to read the paper individually and then meet to discuss the paper, and decide which parts of the paper to emphasize and cover in the presentation. All members of the group should speak during the presentation.
    • Each group's presentation will be a maximum of 10 minutes long, with questions. Plan to speak for 7-8 minutes at most, and you should rehearse your talk.
    • Your audience is the other students in the course. You should present the material in a way that will be understandable by the class, without oversimplifying the paper content. Your goal is to clearly summarize the problem addressed and key contributions of the research paper, along with your own analysis of the work. Some questions to consider as you prepare your presentation are:
      1. Are the conclusions well-supported by the results provided?
      2. Is the approach and/or strategies employed by the author(s) sensible in your opinion? Why or why not?
      3. Are the contributions incremental (i.e. minor improvements for existing techniques), or represent a radical departure or new methdology?
    • Grading: Groups will receive a grade out of 20 points (10 points for content, and 10 points for clarity).
  • Read Ch. 2 of the Duin text (available online through MyCourses)

Week 2

  • Assignment 1 has been posted on MyCourses (due Mon. Feb. 8th)
  • Prof. Zanibbi's office hours have been updated (see above).

Week 1

  • Reading: Read the Elements of Statistical Learning (link available through MyCourses), Chs. 1 and 2.1-2.4 before class on Friday.
  • Prof. Zanibbi's office hours have been updated to Mondays 1-2pm, and Fridays 1-2pm and 3-4pm. Students can also try setting up an appointment over email if they have a conflict and need to see me out-of-class.
  • The first assignment will be given out Monday of Week 2.
  • The course syllabus and schedule are now available.
  • Slides introducing pattern recognition may be found on MyCourses.