Visual deconstruction

Recognizing articulated objects

Tyng Luh Liu, Davi Geiger

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

Abstract

We propose a deconstruction framework to recognize and find articulated objects. In particular we are interested in human arm and leg articulations. The deconstruction view of recognition naturally decomposes the problem of finding an object in an image, into the one of (i) extracting key features in an image, (ii) detecting key points in the models, (iii) segmenting an image, and (iv) comparing shapes. All of these subproblems can not be resolved independently. Together, they reconstruct the object in the image. We briefly address (i) and (ii) to focus on solving together shape similarity and segmentation, combining top-down & bottom-up algorithms. We show that the visual deconstruction approach is derived as an optimization for a Bayesian-Information theory, and that the whole process is naturally generated by the guaranteed Dijkstra optimization algorithm.

Original languageEnglish (US)
Title of host publicationEnergy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings
PublisherSpringer Verlag
Pages295-309
Number of pages15
Volume1223
ISBN (Print)3540629092, 9783540629092
DOIs
StatePublished - 1997
EventInternational Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition, EMMCVPR 1997 - Venice, Italy
Duration: May 21 1997May 23 1997

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1223
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

OtherInternational Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition, EMMCVPR 1997
CountryItaly
CityVenice
Period5/21/975/23/97

Fingerprint

Information theory
Dijkstra Algorithm
Information Theory
Bottom-up
Optimization Algorithm
Segmentation
Decompose
Vision
Object
Optimization
Model
Human
Similarity
Framework

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Liu, T. L., & Geiger, D. (1997). Visual deconstruction: Recognizing articulated objects. In Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings (Vol. 1223, pp. 295-309). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1223). Springer Verlag. https://doi.org/10.1007/3-540-62909-2_87

Visual deconstruction : Recognizing articulated objects. / Liu, Tyng Luh; Geiger, Davi.

Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings. Vol. 1223 Springer Verlag, 1997. p. 295-309 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1223).

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

Liu, TL & Geiger, D 1997, Visual deconstruction: Recognizing articulated objects. in Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings. vol. 1223, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1223, Springer Verlag, pp. 295-309, International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition, EMMCVPR 1997, Venice, Italy, 5/21/97. https://doi.org/10.1007/3-540-62909-2_87
Liu TL, Geiger D. Visual deconstruction: Recognizing articulated objects. In Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings. Vol. 1223. Springer Verlag. 1997. p. 295-309. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/3-540-62909-2_87
Liu, Tyng Luh ; Geiger, Davi. / Visual deconstruction : Recognizing articulated objects. Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings. Vol. 1223 Springer Verlag, 1997. pp. 295-309 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{48be3c7056744eeab5b15b7ab10949f5,
title = "Visual deconstruction: Recognizing articulated objects",
abstract = "We propose a deconstruction framework to recognize and find articulated objects. In particular we are interested in human arm and leg articulations. The deconstruction view of recognition naturally decomposes the problem of finding an object in an image, into the one of (i) extracting key features in an image, (ii) detecting key points in the models, (iii) segmenting an image, and (iv) comparing shapes. All of these subproblems can not be resolved independently. Together, they reconstruct the object in the image. We briefly address (i) and (ii) to focus on solving together shape similarity and segmentation, combining top-down & bottom-up algorithms. We show that the visual deconstruction approach is derived as an optimization for a Bayesian-Information theory, and that the whole process is naturally generated by the guaranteed Dijkstra optimization algorithm.",
author = "Liu, {Tyng Luh} and Davi Geiger",
year = "1997",
doi = "10.1007/3-540-62909-2_87",
language = "English (US)",
isbn = "3540629092",
volume = "1223",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "295--309",
booktitle = "Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings",

}

TY - GEN

T1 - Visual deconstruction

T2 - Recognizing articulated objects

AU - Liu, Tyng Luh

AU - Geiger, Davi

PY - 1997

Y1 - 1997

N2 - We propose a deconstruction framework to recognize and find articulated objects. In particular we are interested in human arm and leg articulations. The deconstruction view of recognition naturally decomposes the problem of finding an object in an image, into the one of (i) extracting key features in an image, (ii) detecting key points in the models, (iii) segmenting an image, and (iv) comparing shapes. All of these subproblems can not be resolved independently. Together, they reconstruct the object in the image. We briefly address (i) and (ii) to focus on solving together shape similarity and segmentation, combining top-down & bottom-up algorithms. We show that the visual deconstruction approach is derived as an optimization for a Bayesian-Information theory, and that the whole process is naturally generated by the guaranteed Dijkstra optimization algorithm.

AB - We propose a deconstruction framework to recognize and find articulated objects. In particular we are interested in human arm and leg articulations. The deconstruction view of recognition naturally decomposes the problem of finding an object in an image, into the one of (i) extracting key features in an image, (ii) detecting key points in the models, (iii) segmenting an image, and (iv) comparing shapes. All of these subproblems can not be resolved independently. Together, they reconstruct the object in the image. We briefly address (i) and (ii) to focus on solving together shape similarity and segmentation, combining top-down & bottom-up algorithms. We show that the visual deconstruction approach is derived as an optimization for a Bayesian-Information theory, and that the whole process is naturally generated by the guaranteed Dijkstra optimization algorithm.

UR - http://www.scopus.com/inward/record.url?scp=84958624964&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84958624964&partnerID=8YFLogxK

U2 - 10.1007/3-540-62909-2_87

DO - 10.1007/3-540-62909-2_87

M3 - Conference contribution

SN - 3540629092

SN - 9783540629092

VL - 1223

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

SP - 295

EP - 309

BT - Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings

PB - Springer Verlag

ER -