CSE 690 Internet Vision

jabgoldfishAI and Robotics

Oct 19, 2013 (3 years and 5 months ago)


CSE 690 Internet Vision

Organizational Meeting

Tamara Berg

Assistant Professor

SUNY Stony Brook

Course Information

Instructor: Tamara Berg

Email: tlberg@cs.sunysb.edu

Office: 1411 Computer Science

Webpage: http://tamaraberg.com/

Course: Time/Location TBD

Office Hours: TBD

Course Webpage: http://tamaraberg.com/teaching/Fall_08

About Me

First year as a professor

PhD from University of California, Berkeley 2007

Research Scientist at Yahoo! Research 2007


First time teaching so I’ll be learning as we go along
with all of you!


Digital Media, Computer Vision, Natural Language
Processing, Web Scale projects, Integrating Words &

About You?





Any related background in Computer Vision,
Machine Learning, Graphics?

Automatic Image Completion

James Hays, Alexei A. Efros. Scene Completion Using Millions of Photographs.

Face Transfer

"Face Swapping: Automatically Replacing Faces in Photographs,” D. Bitouk, N. Kumar, S.
Dhillon, P. Belhumeur, S. K. Nayar, Siggraph 2008

Photo Tourism

Noah Snavely, Steven M. Seitz, Richard Szeliski, "Photo tourism: Exploring photo
collections in 3D,” SIGGRAPH 2006

Automatic Photo Pop

D. Hoiem, A.A. Efros, and M. Hebert, "Automatic Photo Pop

Course Organization


Student Paper Presentations

Course Project

Your Responsibilities

Attend lectures

Read assigned papers before each class and
write down at least a few questions

Participate in discussions!

Student Paper Presentations

Course project

Have fun! New and exciting area of
research with broad industrial applications.

Lecture Topics

Visual and Multi
Media Data

Computational Photography

Text Based Retrieval

Exemplar Based Retrieval

Photo Quality for Retrieval

Combining Words & Pictures


Objects, People & Events

The role of Social Networks & Human Interaction

Including cutting edge research in: Computer Vision, Graphics,
Multimedia, Information Retrieval, Machine Learning, HCI

Student Paper Presentations

You be the professor for a day

Prepare a presentation and lead a discussion
on a few of the assigned research papers
including discussion questions, pros/cons etc

Present a demo of the system if one is

Feel free to run your presentation by me in
office hours beforehand


Course will be project focused

You will pick a project idea near the beginning of the semester

I can
help suggest or refine project ideas in office hours.

Projects can be implementing/modifying one of the papers we discuss
or original research

There will be a few status updates over the semester where you will
briefly describe progress and we can provide suggestions

Projects can be completed alone or in pairs

Hopefully some of these projects can be submitted to a conference!

Worried about the material?

I will present a summary of useful knowledge,
algorithms and related techniques near the
beginning of the semester.

This should enable you to understand most of
the research papers we will read.

If you are still worried, feel free to come see
me in office hours.


2 classes per week, 1:20 long

Discuss 1
2 papers per class

1 class per week, 2:30 long

Discuss ~3 papers per class

Informal discussions, “brown bag” seminar?

Lecture Times?

For next class


Please Read (links will be posted on course webpage):

80 million tiny images: a large dataset for non
parametric object and scene recognition

Course Information

Instructor: Tamara Berg

Email: tlberg@cs.sunysb.edu

Office: 1411 Computer Science

Webpage: http://tamaraberg.com/

Course Time: TBD

Course Location: Will be posted on webpage

Office Hours: TBD

Course Webpage: http://tamaraberg.com/teaching/Fall_08