|
Instructor:
Prof. Sinisa Todorovic sinisa at eecs oregonstate edu 2107 Kelley Engineering Center Classes: TR 8:30-9:50am, Milam Hall 234 Office hours: W 4-5pm, or by appointment
|
 |
| LECTURE NOTES: |
| Date | Topics and Recommended Literature | Slides |
| 09/30 | Introduction: Administrative stuff; What is computer vision; Overview of fundamental problems and popular applications; Relationship of computer vision with other research fields; Course overview |  |
| 10/02 | Big
picture -- Image formation; Properties of 2D objects; From image
structure to image understanding; David
Marr: Cognition
as computation; David
Marr's paradigm: Sequential processing of the visual information; Gestalt laws
(Forsyth & Ponce pp. 304-309); Barrow
and Tenenbaum: Intrinsic images; Biederman:
Recognition-by-components |
|
| 10/07 | Big picture (continued); Image features; Color (Forsyth & Ponce pp. 97-132); Edges (Forsyth & Ponce pp. 165-188); Image filtering (Forsyth & Ponce pp. 135-164); Matlab warm-up | |
| 10/09 | Image features; Interest points and keypoint descriptors; Wavelets; Homework 1 (due 10/21) | |
| 10/14 | Pinhole
camera (Forsyth & Ponce pp. 3-19); Camera parameters and
perspective projection (Forsyth & Ponce pp. 20-37); MATLAB functions
for multiple view geometry; |
|
| 10/16 |
Camera parameters and perspective projection (continued); Camera calibration (Forsyth & Ponce pp. 38-54); | |
| 10/21 HW1
|
Epipolar geometry and weak calibration (Forsyth & Ponce pp. 215-233); Homework 2 (due 10/30) | |
| 10/23 |
Course project guidelines (Project proposal due 11/04); | |
| 10/28 | 2D
Homography; RANSAC
(Forsyth
& Ponce pp. 346-351); Voronoi
diagram; Low-level
segmentation; |
|
| 10/30 HW2
|
Clustering
-- Normalized
Cuts (Forsyth & Ponce pp. 313-328); Relaxation Labeling;
Multiscale image segmentation: Scale-space
and Integrating edge
and region detection; |
|
| 11/04 Proj
|
Meanshift; MRFs (Tutorial by Bouman, Tutorial by Perez); | |
| 11/06 |
MRFs (Tutorial by Bouman, Tutorial by Perez); | |
| 11/11 |
MRFs (continued); 2D Object recognition; | |
| 11/13 | 2D Object recognition; Pictorial Structures (Fischler & Elshlanger) and demo; Generative-model based object categorization (Constellation, Latent topic models); Datasets and competitions | |
| 11/18 | Hierarchical object representations; Composition systems; Taxonomy; Connected Segmentation Tree and Taxonomy; Hierarchy of edges Statistical image grammars; Probabilistic grammar Markov models | |
| 11/20 | Face recognition; Scene understanding; Multimodality of words and images; Image retrieval (Forsyth & Ponce pp. 599-619) | |
| 11/25 | Texture
(Forsyth & Ponce pp. 189-212, Survey by Tuceryan & Jain);
Object tracking (Forsyth
&
Ponce pp. 374-398); People
tracking
and human
activity recognition |
|
| 11/27 | Thanksgiving
Holiday |
|
| 12/02 | Presentations |
|
| 12/04 |
Presentations |
|