BCS Course Materials

Description | Syllabus | Archives |

BCS 220

The Intelligent Eye

Instructor: David Knill

Office: 275 Meliora

Phone: 275-4597

e-mail: knill@cvs.rochester.edu

Office hours:

T TH 1:00 — 2:30

Catalog description:

Aristotle described visual perception as "to know what is where by looking". This common-sense description of what it means to see reflects our phenomenal experience with the ease and simplicity of perception. This apparent simplicity is the first of many illusions that students will encounter in the class. Computers have achieved grand master status in chess, yet still cannot solve vision problems that young children handle with ease every second of their waking day. The course will provide an inter-disciplinary view of modern research into how the human brain solves the problems involved in perception, including how we perceive the three dimensional structure of the world, how we recognize objects and how visual information is used to control action in the world. Students will read contemporary research papers that will expose them to the science that has forced us to re-conceptualize, sometimes in non-intuitive ways, how vision works. Through classroom discussion and a series of critical essays, students will explore and analyze the questions and debates that define contemporary perceptual science.

Reading:

Reading for the course will include book chapters and research articles from the vision literature. A course pack will be available within a week of the beginning of class. This will contain copies of the articles covered in the class. Copies of book chapters will be handed out in class.

 

Organization:

Basic structure: The course will be run as a cross between a lecture and a seminar. The main goal of the course is to critically consider contemporary research in vision; thus, many of the class meetings will focus on discussions of research papers. A certain amount of content material (e.g. how cells in primary visual cortex work) is needed to understand much of the material, so interspersed with the paper discussions will be lectures based on readings from textbooks.

Student participation: By its nature, the success of the class depends largely on your involvement. In order to make the course work, you will have to come to class prepared — having done the reading and writing assignments for the class. I expect everyone to participate in the class. This means that attendance is mandatory (excepting absences due to valid reasons such as sickness) and I expect all students to participate in discussion. This will make the class fun for all of us, including myself.

Regular Assignments: Course work consists primarily of reading chapters from books and articles from the scientific literature, and regular, small writing assignments. You will be doing two types of writing assignments -

• Thought questions — You will write brief answers to a small number (1-3) of thought questions about the reading for each class. I will hand these out the class before they are due. The questions are designed to push you to think critically about the reading prior to coming to class. No thought questions will be due the days on which you turn in reflection papers (see below).

• Reflection papers — You will write six 3-page reflection papers during the course of the term. These will be due at the end of each of the first six main segments of the class. Each reflection paper will address a question concerning the topic covered in the corresponding segment of the class. In answering the question, you should integrate information from the material covered in the readings and the classroom discussion/lecture.

Grading and revision policy:

Thought questions and reflection papers will be graded on a simple contract basis — 1 point for satisfactory, 0 points for unsatisfactory. In order to receive a point, an assignment must meet the following criteria

Contract for thought questions

.• A coherent first attempt at answering the question must be mailed to me over electronic mail before the beginning of class. You should also bring a copy of your answers to the class. Answers should be written so that they could be understood by other students in the class. No credit will be given, or revisions allowed, for incoherent initial answers.

• An answer is unsatisfactory if it reflects a misunderstanding of the reading material or is built on faulty logic.

• Revision policy - You have the opportunity to revise any answers that are unsatisfactory and turn in the revised answers one class period after the initial due date. Revised answers must be e-mailed to me by midnight on this later date in order to be considered for a point. I will be available to meet with you if you need help with any questions.

• 1 point is given for each assignment as a whole. Any answer deemed unsatisfactory renders the entire assignment unsatisfactory. Revisions need only address the unsatisfactory answer(s).

• I will e-mail you a list of the answers that are unsatisfactory before the morning after you have turned them in.

Contract for reflection papers

• A reasonable draft of the reflection paper must be turned in, typewritten, at the beginning of the class on which it is due. The paper turned in should be written at the level of a second draft of a paper. Rough drafts, for example, freewritten lists of ideas used to jumpstart your paper, will not be accepted.

• To be satisfactory, the paper should contain a well-structured and coherent argument that integrates the material from the appropriate class segment to address the question posed.

• Revision policy - You have the opportunity to revise a paper that is unsatisfactory and turn in the revised paper one week after the paper is returned to you. Revised papers must be turned in at the beginning of class at this later date in order to be considered for a point. I will be available to meet with you if you need help with any questions.

Final project

Grant proposal

The final project will be a ten page grant proposal, in which you define a research question (or questions) focused on a particular problem in vision. The grant proposal should include a 1-page introductory statement of the research question(s) and its significance, a 4 page literature review and a 5 page description of the proposed research plan. I will hand out a more detailed set of writing guidelines at the midpoint of the term. Prior to beginning the final project, you will turn in a 2 page pre-proposal outlining your ideas. I will then meet with you individually to help refine the topic and scope of your paper.

After turning in the paper on April 18 and getting feedback from the class (from oral presentations) and myself, you will revise your paper and turn in a final draft by May 8.

Oral presentation

As part of the final project, you will give an oral presentation (30 minutes) describing your proposed research to the class. I will hand out guidelines for the oral presentations several weeks in advance of their start (April 18). In preparation for your in-class presentation, I will meet with each of you individually for a practice run to give you feedback.

Important dates for the final project

March 19: Pre-proposal due

March 19 — March 26: Meetings with students to discuss their projects

April 18: Final papers due

April 18 — May 1: Student presentations

May 8 — revised final projects due

 

Grading

Thought questions: 25%

Reflection papers: 25%

Final project: 25%

In class participation: 25% (10 of these 25 points will be given for simple attendance. The remaining 15 points will be based on the degree to which you participate in classroom discussion)

Attendance

I expect students to attend all classes. 2 points (out of 100) will be deducted from your final grade for every unexcused absence.

Excused absences: Excused, unplanned, absences include sickness and family emergencies. These require appropriate documentation (from Student Health, a hospital, etc.). If you know you have to miss a class for some other reason (e.g. a University sponsored activity or a job interview), you must inform me beforehand and show reasonable documentation. I will not accept any reason for an absence if you knew of it prior to class and did not discuss it with me in advance.

 

COURSE OUTLINE

INTRODUCTION AND OVERVIEW OF APPROACH

Jan. 17 Introduction

Reading due: None

Homework due: None

Jan. 22 Approaches to vision

Readings: Marr, pp. 3 — 38.

Homework due: Thought questions

UNDERSTANDING IMAGE FORMATION AND CODING

Jan. 24 Sensing the optic array

Reading due: Bruce, Green and Georgeson, pp. 3 - 24

Homework due: Thought questions

Jan. 29 Detection and coding

Reading due: Cornsweet, Chapters 2 and 4

Homework due: 1st Reflection paper

SOME SPECIFIC PROBLEMS — FROM COMPUTATIONAL THEORY TO NEURAL PROCESSING (FILTER MODELS)

Jan. 31 Cortical coding, filtering, multi-scale representations

Reading due: Bruce, et. al., 30 — 41, 48 — 56.

Homework due: Thought questions

Feb. 5 Texture segmentation

Reading due: Bergen (1991)

Homework due: Thought questions

Feb. 7 Stereopsis — computational theory

Reading due: Bruce, et. al., 137 — 147, 234 — 235, Marr, 111 - 122

Homework due: Thought questions

Feb. 12 Stereopsis — mechanism

Reading due: Howard and Rogers, 132 - 141

Homework due: 2nd Reflection paper

SURFACE PERCEPTION: CUE INTEGRATION

Feb. 14 Depth perception: cue integration I

Reading due: Johnston, Cumming and Parker (1993)

Homework due: Thought questions

Feb. 19 Cue integration lab

Feb. 21 Lab wrap-up

Reading due: None

Homework due: Data analysis, Thought questions

Feb. 26 Cue integration wrap-up — Bayesian cue integration

Reading due: Saunders and Knill (2001)

Homework due: 3rd Reflection paper

OBJECT RECOGNITION

Feb. 28 Structural models

Reading due: Biederman (1987), Biederman and Gerhardstein (1993)

Homework: Thought questions

 

March 5 View-based models

Reading due: Poggio and Edelman(1990), Bulthoff and Edelman (1992)

Homework: Thought questions

March 7 The debate

Reading due: Tarr, et. al. (1998), Biederman and Barr (1999)

Homework: 4th Reflection paper

March 12, 14 Spring Break

March 19 Face recognition — psychophysics

Reading due: TBD

Homework due: Thought questions

** Project proposals due

March 21 Face recognition — neurophysiology, fMRI

Reading due: TBD

Homework due: Thought questions

VISUAL CONTROL OF ACTION

March 26, 28 OFF

April 2 Optic flow

Reading due: Optic flow handout, Bruce, et. al., pp. 281 — 311

Homework due: Thought questions

April 4 Braking,

Reading due: Yilmaz and Warren (1995)

Homework due: Thought questions

April 9 Direction of heading

Reading due: Crowell and Banks (1994)

Homework due: 5th Reflection paper

MEASURING MOTION

April 11 Motion coding, stage 1: Aperture problem, MT, MST

Reading due: Movshon and Newsome (1996)

Homework due: Thought questions

April 13 Motion coding, direction of heading

Reading due: Sheroy, et. al. (1999)

Homework due: Thought questions

April 18 Student presentations

** First draft of grant proposal due

April 23 Student presentations

April 25 Student presentations

April 30 Student presentations

May 1 Student presentations

** May 8 Final drafts of projects due

BCS 220 Bibliography

Marr, D. (1982) Vision, W. H. Freeman and Co., New York.

Bruce, V; Green, P. R. and Georgeson, M. A. ( 1996) Visual Perception: Physiology, Psychology and Ecology, Psychology Press, East Sussex, UK.

Cornsweet, T. N. (1970) Visual Perception, Academic Press, New York

Bergen, J. R. (1991) Theories of visual texture perception, In D. M. Regan (ed.) Spatial Vision (Vol. 10 of Vision and Visual Dysfunction) (114-134),, MacMillan, New York.

Howard, I. P. and Rogers, B. J. (1995) Binocular Vision and Stereopsis , Oxford University Press, New York.

Johnston, E. B., Cumming, B. G. and Parker, A. J. (1993) Integration of depth modules: Stereopsis and texture, Vision Research, 33, 813 — 826.

Saunders, J. and Knill, D. C. (2001) Perception of 3D surface orientation from skew symmetry, Vision Research, 41 (24), 3163 — 3185.

Biederman, I. (1987) Recognition-by-components: A theory of human image understanding. Psychological Review 94, 115-147.

Biederman, I. & Gerhardstein, P. C. (1993) Recognizing depth-rotated objects: Evidence and conditions for three-dimensional viewpoint invariance. Journal of Experimental Psychology: Human Perception and Performance 19, 1162-1182.

Poggio T. S. and Edelman, S. (1990) Nature, 343, 263-266.

Bulthoff, H. B. and Edelman, S. (1992) Psychophysical support for a two-dimensional view interpolation theory of object recognition, Proceedings of the Nat'l Acad. of Sciences, 89, 60-64.

Tarr, M. J., Williams, P., Hayward, W. G. and Gauthier, I. (1998) Three dimensional object recognition is viewpoint-dependent, Nature Neuroscience, 1, 275-277.

Biederman, I. and Bar, M. (1999) One shot viewpoint invariance in matching novel objects, Vision Research, 39, 288555-2899.

Yilmaz, E. H. and Warren, W. H. (1995) Visual control of braking — A test of the tau-hypothesis, J. Exp. Psych: Human Perc. and Perf., 21 (5), 996 — 1014.

Royden, C. S., Crowell, J. A. and Banks, M. S. (1994) estimating heading during eye movements, Vision Research, 34, 3197 — 3214.

Movshon JA, Newsome WT (1996) Visual response properties of striate cortical neurons projecting to area MT in macaque monkeys, J. of Neuroscience, 16, 7733-7741.

Shenoy KV, Bradley DC, Andersen RA (1999) Influence of gaze rotation on the visual response of primate MSTd neurons, J. of Neurophysiology, 81 (6): 2764-2786.


Last modified: 5/7/2002
questions and comments
about this site