WMVL banner
WMVL banner Penn State Mark WMVL banner
WMVL banner

Home
News
People
Research
Publications
Courses
Seminars
Data &Code
Contact Us
Calendar
Other Links

CVPR 2013: Symmetry Detection from Real World Images -- A Competition

  symr2  frieze  06038_Sinosteel_Facade

Symmetry is a pervasive phenomenon present in all forms and scales in natural and man-made environments. It is not surprising, therefore, that humans, animals and insects have evolved an innate ability to perceive and take advantage of symmetry. What IS surprising is that perception and recognition of symmetry have yet to be fully explored in machine intelligence, in particular computer vision. Despite an understanding of how the concept of repeated patterns is generalized by the mathematics of group theory, and despite attempts over four decades to design algorithms that seek symmetry from digital data, there are very few effective computational tools for automated symmetry analysis available today. The goal of this competition is to benchmark state of the art symmetry detection algorithms (previously published and new algorithms) on real images.

0. Keynote Speaker:


Qasim Zaidi
Distinguished Professor
(SUNY)

Parallel model for the salience of mirror symmetry in natural patterns

Mirror symmetry detection is considered extremely rapid, based on experiments with random noise.  Symmetry detection in natural settings, however, is accomplished against structured backgrounds. We measured temporal thresholds for detecting the mirror symmetry axis in patterns assembled from 101 natural textures. Thresholds ranged from 0.028 to 0.568 sec indicating a wide range of salience (1/Threshold). For estimating symmetric energy, we used pairs of mirror-symmetric cortex-like filters, connected with AND junctions, in a filter-prune-filter-select model. The model easily identified the axis of symmetry for all patterns, but symmetry magnitude quantified at this axis correlated weakly with salience...

1. Co-Chairs:

yanxi
Yanxi Liu
 (PSU)

lucVanGool
Luc Van Gool (ETHZ & Univ. of Leuven)

2. Organizing Committee:

lsk
Seungkyu Lee
(SamsungResearch)

jc
Jingchen Liu
(PSU)

minwoo
Minwoo Park
(ObjectVideo)

gang
Gang Zheng
(PSU)

george_slota
George Slota
(PSU)

wzh2
Zhaohui Wu
(PSU)

3. Advisory Committee:

jacob
Jacob Feldman    (Rutgers)

hartley

Richard Hartley

(ANU)

kanade_takeo

Takeo Kanade

(CMU)

malik

Jithendra Malik

(U.C. Berkeley)

schatt

Doris Schattschneider (Moravian College)

Marjorie

Marjorie Senechal

(Smith College)

tyler
Christopher Tyler (SKBIC)

younesPicture
Laurent Younes (Johns Hopkins University)

alan

Alan Yuille

(UCLA)

fellows-2008-az

Andrew Zisserman

(Oxford)

                       

 

 

Schedule (06/23/2013)

time

topic

speaker

type

10:15-10:45

CVPR break time

10:45-10:55

 

 

An INTRODUCTION (from a historical perspective) of the symmetry detection competition

 

Description: yanxi

Dr. Y. Liu

10:55-11:10

 

 

A summary of the Reflection Symmetry Detection (submission and evaluation)

 

 

Description: jc
J. Liu

11:10-11:25

Recognition of Symmetry Structure by Use of Gestalt Algebra

E. Michaelsen

reflection

11:25-11:40

 

Detection of Mirror-Symmetric Image Patches

 

VPatraucean

V. Patraucean

reflection

11:40-11:55

Multi-Scale Kernel Operators for Reflection and Rotation Symmetry

A. Petrosino

reflection

12:00-13:30

CVPR lunch

13:30-13:40

 

Summary on Rotation Symmetry Detection (submission and evaluation)

 

Description: lsk

Dr. S. Lee

13:40-13:50

Multi-Scale Kernel Operators for Reflection and Rotation Symmetry

A. Petrosino

rotation

13:55-14:10

Recognition of Symmetry Structure by Use of Gestalt Algebra

E. Michaelsen

translation

14:10-14:25

 

Translation Symmetry Detection: A Repetitive Pattern Analysis Approach

CaiYunliang
Y. Cai

translation

14:25-15:15

 

 

Keynote Speech:

Parallel model for the salience of mirror symmetry in natural patterns

 

Qasim Zaidi

15:35-15:55

CVPR break time

15:55-16:00

Awards for the winners

 

 

16:00-17:00

 

 

 

 

PANEL discussion:

What should be the Gold-standard for evaluating "Symmetry Detection from Real Images" algorithms in computer vision: human perception or mathematics?

 

4. Results and Paper Submissions:

            We are releasing the training datasets containing images and hand-labeled ground truth, representative of the types of symmetry that will be found in the test set. After the release of the testing dataset (mid April), competitors are required to submit results (with a same format as our ground truth) for evaluation, along with an associated paper describing the algorithm(s) used. We encourage new symmetry detection algorithm or evaluating published algorithms on our benchmark dataset. A special issue for the competition results in the Journal of Machine Vision and Applications is planned.

 

5. Important Dates:  

Training Data Release (with groundtruth labels)       

Feb.01

Testing Data Release (with no groundtruth labels)        

Mar.01

Workshop Paper & Testing Results Submission (extended)

Apr.24

Decision Announcement based on both the paper and the results    

May01

Camera-ready Submission

May07

Workshop

Jun.23

           

6. Important Links:

            Call for Real World Symmetry Images Contribution
            First Symmetry Detection Competition (CVPR 2011)

            PSU Near-Regular Texture Database

            Survey paper:

            -- Yanxi Liu and Hagit Hel-Or and Craig S. Kaplan and Luc Van Gool, Computational Symmetry in Computer Vision and Computer Graphics, Foundations and Trends in Computer Graphics and Vision, Vol.5, Num.1-2, Pages 199, 2010. [pdf]        

7. Data Download:   

Reflection Symmetry

Rotation Symmetry

Translation Symmetry

gt_manmade_15

training set

roteg

training set

transEg

training set

 

 

 

testing set

 

groundtruth

 

baseline alg.

testing set

 

groundtruth

 

baseline alg.

testing set

 

groundtruth

 

baseline alg.

 

submit

 

submit

 

submit

 

8. Submission:

 

            We invite all participants to submit a 4 page workshop paper using the CVPR12 template to describe their algorithms as well as their detection results on the testing dataset in the same format of the groundtruth label we release for the training data.

            W also need a normalized confidence score (between 0 and 1) to be associated with each detected potential symmetry patterns. We recommend participants to include detection results (>=3 per image) even with low confidence scores so that we can get a more complete precision/recall curve for evaluation. 

[example submission]

 

 

Sponsors:

nsf

cvf

ieee_cs

cvpr13logo

 

WMVL banner


Maintained by LPAC webmaster

WMVL banner WMVL banner