CMU Advanced Perception Class, Spring 2004


Tuesday-Thursday 10:30AM-11:50AM NSH 3002



 
Instructor of record: Martial Hebert hebert@ri.cmu.edu

Table of Contents

Class Format

The Advanced Perception course is a graduate class, meeting twise a week to discuss a set of papers covering a specific topic in computer vision and perception. The intend is to look at important work in the field of computer vision with an emphasis on the most current work reported in the most recent conference proceedings and journals.

The instructor will cover background material in class as introduction to each of the topics. Each week, two papers on a particular topic will be assigned. After reading them, your must find a third paper on your own that is relevant to the topic. You need to bring a hard copy of the paper you selected to the instructor and send an email message containing the full citation and one-line summary of the paper.  During class, each of the two assigned papers will be formally presented by one of the students (mapping students to topics will be done in the first week of class). This is expected to be a formal  presentation in front of the class (approximately 40 mins). The presenter will then answer the questions from other students. The presentation will then evolve into a class discussion on the topic covered in the paper. The instructors are responsible for keeping the discussion in a fruitful vein and making sure all students get a chance to participate. The instructors are also responsible for making sure that the important points are touched upon during the discussion, which will sometimes mean asking questions of the class, and for making sure that each paper is covered (which sometimes means cutting off discussion and moving on). It is expected that proof-of-concept implementation of key concepts will be generated in order to illustrate the presentation.
At the end of the class, we will go around the room asking each of you to cite the third paper you have personally chosen for that week, briefly describe it (but in sufficient details to understand the point of the paper), tell us why you picked it (i.e. how does it relate to the topic area and the two assigned papers), and finally whether or not you would recommend that paper for others to read.
Grading is based entirely on the quality of the presentations, analysis of the papers, and discussion.

It is important that students in the class have previously taken 16-720 or a similar intro to Computer Vision.

Computer Vision Resources

There are many places to go to look for computer vision papers, ranging from archival journals to on-line web sites. Here is a list of our favorite sources of material:

Archival Journals

Reference Books

  • "Computer Vision: A Modern Approach" by David Forsyth and Jean Ponce, 2002.
  • "Vision Science", Stephen Palmer, 2000, MIT Press.
  • "The Geometry of Multiple Images" by Olivier Faugeras, MIT Press, 2001.
  • "Multiple View Geometry", Richard Hartley and Andrew Zisserman, Cambridge University Press, 2000.
  • "Introductory Techniques for 3-D Computer Vision" by Emanuele Trucco and Alessandro Verri, Prentice Hall, 1998.
  • "Epipolar Geometry in Stereo, Motion and Object Recognition -- A Unified Approach" by Gang Xu and Zhengyou Zhang, Kluwer Academic Publishers, 1996.
  • "Three-Dimensional Computer Vision -- A Geometric Viewpoint" by Olivier Faugeras, The MIT Press, 1993.
  • "Active Contours" by Andrew Blake and Michael Isard, Springer, 1998.
  • Conference Proceedings

    WWW Resources

    2004 Topics

    This year's selection of papers is devoted to object recognition/detection (in a broad interpretation of the term) since that topic was somewhat neglected in 16-720 and it is an increasingly important area.

    Note that, one some weeks, two papers may be listed for a single presentation. This occurs when two papers are closely related (e.g., one is an earlier version of a later implementation), for example. Also, some weeks include background papers that are intended to facilitate the understanding of the main papers.

    Current Selection of Papers by Week (Selections subject to change)