Articulation-invariant representation of non-planar shapes

Raghuraman Gopalan, Pavan Turaga, Rama Chellappa

Research output: Chapter in Book/Report/Conference proceedingConference contribution

73 Citations (Scopus)

Abstract

Given a set of points corresponding to a 2D projection of a non-planar shape, we would like to obtain a representation invariant to articulations (under no self-occlusions). It is a challenging problem since we need to account for the changes in 2D shape due to 3D articulations, viewpoint variations, as well as the varying effects of imaging process on different regions of the shape due to its non-planarity. By modeling an articulating shape as a combination of approximate convex parts connected by non-convex junctions, we propose to preserve distances between a pair of points by (i) estimating the parts of the shape through approximate convex decomposition, by introducing a robust measure of convexity and (ii) performing part-wise affine normalization by assuming a weak perspective camera model, and then relating the points using the inner distance which is insensitive to planar articulations. We demonstrate the effectiveness of our representation on a dataset with non-planar articulations, and on standard shape retrieval datasets like MPEG-7.

Original languageEnglish (US)
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages286-299
Number of pages14
Volume6313 LNCS
EditionPART 3
DOIs
StatePublished - 2010
Externally publishedYes
Event11th European Conference on Computer Vision, ECCV 2010 - Heraklion, Crete, Greece
Duration: Sep 5 2010Sep 11 2010

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 3
Volume6313 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other11th European Conference on Computer Vision, ECCV 2010
CountryGreece
CityHeraklion, Crete
Period9/5/109/11/10

Fingerprint

Cameras
Decomposition
Imaging techniques
Invariant
Convex Decomposition
MPEG-7
Occlusion
Set of points
Normalization
Convexity
Retrieval
Camera
Imaging
Projection
Modeling
Demonstrate
Model

Keywords

  • articulations
  • convex decomposition
  • Shape representation

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Gopalan, R., Turaga, P., & Chellappa, R. (2010). Articulation-invariant representation of non-planar shapes. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (PART 3 ed., Vol. 6313 LNCS, pp. 286-299). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6313 LNCS, No. PART 3). https://doi.org/10.1007/978-3-642-15558-1_21

Articulation-invariant representation of non-planar shapes. / Gopalan, Raghuraman; Turaga, Pavan; Chellappa, Rama.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 6313 LNCS PART 3. ed. 2010. p. 286-299 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6313 LNCS, No. PART 3).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Gopalan, R, Turaga, P & Chellappa, R 2010, Articulation-invariant representation of non-planar shapes. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 3 edn, vol. 6313 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 3, vol. 6313 LNCS, pp. 286-299, 11th European Conference on Computer Vision, ECCV 2010, Heraklion, Crete, Greece, 9/5/10. https://doi.org/10.1007/978-3-642-15558-1_21
Gopalan R, Turaga P, Chellappa R. Articulation-invariant representation of non-planar shapes. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 3 ed. Vol. 6313 LNCS. 2010. p. 286-299. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3). https://doi.org/10.1007/978-3-642-15558-1_21
Gopalan, Raghuraman ; Turaga, Pavan ; Chellappa, Rama. / Articulation-invariant representation of non-planar shapes. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 6313 LNCS PART 3. ed. 2010. pp. 286-299 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3).
@inproceedings{bae4c44227204afb81d50e33b1e0a0a8,
title = "Articulation-invariant representation of non-planar shapes",
abstract = "Given a set of points corresponding to a 2D projection of a non-planar shape, we would like to obtain a representation invariant to articulations (under no self-occlusions). It is a challenging problem since we need to account for the changes in 2D shape due to 3D articulations, viewpoint variations, as well as the varying effects of imaging process on different regions of the shape due to its non-planarity. By modeling an articulating shape as a combination of approximate convex parts connected by non-convex junctions, we propose to preserve distances between a pair of points by (i) estimating the parts of the shape through approximate convex decomposition, by introducing a robust measure of convexity and (ii) performing part-wise affine normalization by assuming a weak perspective camera model, and then relating the points using the inner distance which is insensitive to planar articulations. We demonstrate the effectiveness of our representation on a dataset with non-planar articulations, and on standard shape retrieval datasets like MPEG-7.",
keywords = "articulations, convex decomposition, Shape representation",
author = "Raghuraman Gopalan and Pavan Turaga and Rama Chellappa",
year = "2010",
doi = "10.1007/978-3-642-15558-1_21",
language = "English (US)",
isbn = "364215557X",
volume = "6313 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
number = "PART 3",
pages = "286--299",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
edition = "PART 3",

}

TY - GEN

T1 - Articulation-invariant representation of non-planar shapes

AU - Gopalan, Raghuraman

AU - Turaga, Pavan

AU - Chellappa, Rama

PY - 2010

Y1 - 2010

N2 - Given a set of points corresponding to a 2D projection of a non-planar shape, we would like to obtain a representation invariant to articulations (under no self-occlusions). It is a challenging problem since we need to account for the changes in 2D shape due to 3D articulations, viewpoint variations, as well as the varying effects of imaging process on different regions of the shape due to its non-planarity. By modeling an articulating shape as a combination of approximate convex parts connected by non-convex junctions, we propose to preserve distances between a pair of points by (i) estimating the parts of the shape through approximate convex decomposition, by introducing a robust measure of convexity and (ii) performing part-wise affine normalization by assuming a weak perspective camera model, and then relating the points using the inner distance which is insensitive to planar articulations. We demonstrate the effectiveness of our representation on a dataset with non-planar articulations, and on standard shape retrieval datasets like MPEG-7.

AB - Given a set of points corresponding to a 2D projection of a non-planar shape, we would like to obtain a representation invariant to articulations (under no self-occlusions). It is a challenging problem since we need to account for the changes in 2D shape due to 3D articulations, viewpoint variations, as well as the varying effects of imaging process on different regions of the shape due to its non-planarity. By modeling an articulating shape as a combination of approximate convex parts connected by non-convex junctions, we propose to preserve distances between a pair of points by (i) estimating the parts of the shape through approximate convex decomposition, by introducing a robust measure of convexity and (ii) performing part-wise affine normalization by assuming a weak perspective camera model, and then relating the points using the inner distance which is insensitive to planar articulations. We demonstrate the effectiveness of our representation on a dataset with non-planar articulations, and on standard shape retrieval datasets like MPEG-7.

KW - articulations

KW - convex decomposition

KW - Shape representation

UR - http://www.scopus.com/inward/record.url?scp=78149335626&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78149335626&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-15558-1_21

DO - 10.1007/978-3-642-15558-1_21

M3 - Conference contribution

SN - 364215557X

SN - 9783642155574

VL - 6313 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 286

EP - 299

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -