Course Organization

1
Course Organization FormatVisualization
IIMSIM 842, CS 795/895

• Instructor
• Jessica Crouch

2
Visualization

• A picture is worth a thousand words
• Visual communication can be
• very effective
• very efficient if done well
• Visual communication can be
• worthless if done poorly

3
Some visualizations are more appealing and
informative than others
4
Course Content

• The question
• How can we design algorithms that produce
  informative, intuitive, graphical representations
  of data?
• No single answer exists (obviously)
• Best design depends on
• Type of information
• What the user needs to learn from the
  visualization
• Efficiency considerations

5
Assumption / Pre-Requisite

• Everyone here knows a bit about graphics and has
  some experience with graphics programming
  (OpenGL)
• What is the difference between graphics and
  visualization?

6
Assumption / Pre-Requisite

• Well focus on designing visualizations and well
  assume you have enough background in graphics to
  implement a design.
• Background knowledge of graphics is important
  because issues of computational efficiency must
  be considered.
• We could dream up NP-Complete algorithms that
  would produce lovely visualizations but would be
  useless from a practical standpoint.
• If a visualization needs to be interactive,
  understanding how the graphics pipeline works is
  especially vital.

7
Types of Information

• How should this course be organized?
• Could categorize according to application area
• Medical
• Networks
• Oceanographic
• Astronomical
• Transportation
• Structural engineering
• Fluid mechanics
• Etc.

8
Types of Information

• Often it is more useful to categorize based on
  fundamental characteristics of the data
• Dimensionality
• Inherent spatial coordinates?
• Time varying?
• Scalar, vector, tensor properties?
• Sparse vs. dense
• Discrete vs. continuous

9
Types of Information

• Visualization algorithms and techniques
  generalize across application areas
• Ex Scalar fields can be visualized using color
  mapping
• Works for temperature distribution over surface
  of the earth, for distribution of radiation over
  a slice of the brain, and for lots of other
  applications

10
Course Organization

• Visualization topics are roughly grouped by data
  type (with a couple exceptions)
• Syllabus tentative schedule are online
• Assigned readings are online too
• Were reading research papers rather than a
  textbook
• Most classes will consist of a presentation and
  discussion of 2 research papers

11
Course Format

• This is a Ph.D. level course
• No higher level courses in visualization are
  offered at ODU
• Objective is to familiarize you with up-to-date
  visualization research
• Focus will be on current research
• Will provide an opportunity for you to
• Learn what is currently going on in the field of
  visualization
• Practice presenting research (even if someone
  elses)
• You need this for conferences, defense, jobs,
  etc.
• Critically evaluating research work
• Youll have to review other peoples papers and
  grant proposals, and youll need to know what to
  expect when other people review your work

12
Paper presentations

• Each student will present two papers during the
  semester
• First task peruse the papers online and
  prioritize them according to which ones you would
  prefer to present
• Email me (jrcrouch_at_cs.odu.edu) your preferences
  by Sunday night. In your email represent each
  paper using the letter labels given on the
  schedule.
• On Monday I will make assignments satisfying as
  many preferences as possible and post the
  presentation schedule online.
• First two presenters then have 9 days to prepare.

13
Course Format

• Look at the website
• www.cs.odu.edu/jrcrouch/courses/msim842-s07
• Write down username, password.

14
How to prepare your presentations

• Read the paper
• Read background material to clarify anything in
  the paper that you dont fully understand
• Look up some of the references from the end of
  the paper
• Use textbooks and other sources as necessary
• Prepare a PowerPoint or pdf presentation and
  lecture to explain the paper
• Send me your presentation file by 5pm Tuesday
  before your Wed. presentation.

15
Presentation organization

• Follow this general outline when preparing your
  presentation
• Problem
• Motivation
• Approach
• Evaluation
• Conclusion
• Questions

16
Problem

• What problem does this work solve?
• Describe how the problem developed
• If possible, tell the problems story
• Ex., New MRI technology was developed that
  reports medically significant tensor data.
  Previous MRI data was scalar valued, so new
  methods for communicating the structure of the
  tensor field in the brain needed to be developed.
  This paper describes a method for visualizing
  two dimensional tensors over the brain volume.
• Give appropriate definitions of any new terms or
  acronyms
• This part should be interesting and attention
  grabbing (while remaining relevant)

17
Motivation

• Why is the problem important?
• Examples of applications that would benefit from
  a solution
• Examples of people who might be able to use a
  good solution
• Would a good solution
• Save lives?
• Save money?
• Help scientists make new discoveries?
• Improve education?
• Tell us why should we care about this work.

18
Approach

• Teach us how this visualization method works
• Provide background information wherever necessary
• Go slow
• Dont skip steps
• Be explicit, give examples
• Include illustrations from the paper and other
  sources
• This is the longest, most detailed part of the
  presentation

19
Evaluation

• How is the performance of this visualization
  measured?
• If a validation experiment was performed, explain
  how it worked inputs, outputs, comparisons
• Summarize the results of the evaluation
• Put in context, given the what competing
  visualization methods might do

20
Conclusion

• In what situations would this method be ideal?
• In what situations would this method perform
  poorly?
• Are the visualizations effective from a human
  perception standpoint?
• Is the method efficient from a computation
  standpoint?
• Is additional work needed? If so, what?
• This is the critical thinking part.

21
Questions

• Bring 3-5 questions for the class
• Ask us questions that require us to thoroughly
  understand the paper in order to answer.
• Ask us questions that make us consider the work
  critically and stimulate discussion.

22
When you are not presenting

• Read the papers carefully before class
• Understand the basic objective and methods used
• Jot down questions regarding anything that is
  unclear
• Make note of what you think the main strengths
  and weaknesses of the work are
• Participate in class discussion of paper
• Provide an honest and constructive evaluation of
  your peers presentations
• See form on course website
• Fill out and give to me (or email me) after class
• I will provide anonymous feedback to presenters

23
Respond to feedback

• No one will present two papers on the same day.
• Please respond to recommendations you receive on
  your first presentation when you prepare your
  second presentation.