Mixing hierarchical contexts for object recognition

Billy Peralta, Alvaro Soto

Resultado de la investigación: Conference contribution

Resumen

Robust category-level object recognition is currently a major goal for the Computer Vision community. Intra-class and pose variations, as well as, background clutter and partial occlusions are some of the main difficulties to achieve this goal. Contextual information in the form of object co-ocurrences and spatial contraints has been successfully applied to reduce the inherent uncertainty of the visual world. Recently, Choi et al. [5] propose the use of a tree-structured graphical model to capture contextual relations among objects. Under this model there is only one possible fixed contextual relation among subsets of objects. In this work we extent Choi et al. approach by using a mixture model to consider the case that contextual relations among objects depend on scene type. Our experiments highlight the advantages of our proposal, showing that the adaptive specialization of contextual relations improves object recognition and object detection performances.

Idioma originalEnglish
Título de la publicación alojadaProgress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 16th Iberoamerican Congress, CIARP 2011, Proceedings
Páginas232-239
Número de páginas8
DOI
EstadoPublished - 28 nov 2011
Evento16th Iberoamerican Congress on Pattern Recognition, CIARP 2011 - Pucon, Chile
Duración: 15 nov 201118 nov 2011

Serie de la publicación

NombreLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen7042 LNCS
ISSN (versión impresa)0302-9743
ISSN (versión digital)1611-3349

Conference

Conference16th Iberoamerican Congress on Pattern Recognition, CIARP 2011
PaísChile
CiudadPucon
Período15/11/1118/11/11

Huella dactilar

Object recognition
Object Recognition
Computer vision
Object Detection
Graphical Models
Clutter
Specialization
Mixture Model
Occlusion
Computer Vision
Uncertainty
Partial
Subset
Context
Object
Experiments
Experiment
Model

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Citar esto

Peralta, B., & Soto, A. (2011). Mixing hierarchical contexts for object recognition. En Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 16th Iberoamerican Congress, CIARP 2011, Proceedings (pp. 232-239). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7042 LNCS). https://doi.org/10.1007/978-3-642-25085-9_27
Peralta, Billy ; Soto, Alvaro. / Mixing hierarchical contexts for object recognition. Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 16th Iberoamerican Congress, CIARP 2011, Proceedings. 2011. pp. 232-239 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{ca78a82ed088490184c5e05c94dab599,
title = "Mixing hierarchical contexts for object recognition",
abstract = "Robust category-level object recognition is currently a major goal for the Computer Vision community. Intra-class and pose variations, as well as, background clutter and partial occlusions are some of the main difficulties to achieve this goal. Contextual information in the form of object co-ocurrences and spatial contraints has been successfully applied to reduce the inherent uncertainty of the visual world. Recently, Choi et al. [5] propose the use of a tree-structured graphical model to capture contextual relations among objects. Under this model there is only one possible fixed contextual relation among subsets of objects. In this work we extent Choi et al. approach by using a mixture model to consider the case that contextual relations among objects depend on scene type. Our experiments highlight the advantages of our proposal, showing that the adaptive specialization of contextual relations improves object recognition and object detection performances.",
author = "Billy Peralta and Alvaro Soto",
year = "2011",
month = "11",
day = "28",
doi = "10.1007/978-3-642-25085-9_27",
language = "English",
isbn = "9783642250842",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "232--239",
booktitle = "Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 16th Iberoamerican Congress, CIARP 2011, Proceedings",

}

Peralta, B & Soto, A 2011, Mixing hierarchical contexts for object recognition. En Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 16th Iberoamerican Congress, CIARP 2011, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7042 LNCS, pp. 232-239, 16th Iberoamerican Congress on Pattern Recognition, CIARP 2011, Pucon, Chile, 15/11/11. https://doi.org/10.1007/978-3-642-25085-9_27

Mixing hierarchical contexts for object recognition. / Peralta, Billy; Soto, Alvaro.

Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 16th Iberoamerican Congress, CIARP 2011, Proceedings. 2011. p. 232-239 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7042 LNCS).

Resultado de la investigación: Conference contribution

TY - GEN

T1 - Mixing hierarchical contexts for object recognition

AU - Peralta, Billy

AU - Soto, Alvaro

PY - 2011/11/28

Y1 - 2011/11/28

N2 - Robust category-level object recognition is currently a major goal for the Computer Vision community. Intra-class and pose variations, as well as, background clutter and partial occlusions are some of the main difficulties to achieve this goal. Contextual information in the form of object co-ocurrences and spatial contraints has been successfully applied to reduce the inherent uncertainty of the visual world. Recently, Choi et al. [5] propose the use of a tree-structured graphical model to capture contextual relations among objects. Under this model there is only one possible fixed contextual relation among subsets of objects. In this work we extent Choi et al. approach by using a mixture model to consider the case that contextual relations among objects depend on scene type. Our experiments highlight the advantages of our proposal, showing that the adaptive specialization of contextual relations improves object recognition and object detection performances.

AB - Robust category-level object recognition is currently a major goal for the Computer Vision community. Intra-class and pose variations, as well as, background clutter and partial occlusions are some of the main difficulties to achieve this goal. Contextual information in the form of object co-ocurrences and spatial contraints has been successfully applied to reduce the inherent uncertainty of the visual world. Recently, Choi et al. [5] propose the use of a tree-structured graphical model to capture contextual relations among objects. Under this model there is only one possible fixed contextual relation among subsets of objects. In this work we extent Choi et al. approach by using a mixture model to consider the case that contextual relations among objects depend on scene type. Our experiments highlight the advantages of our proposal, showing that the adaptive specialization of contextual relations improves object recognition and object detection performances.

UR - http://www.scopus.com/inward/record.url?scp=81855177176&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-25085-9_27

DO - 10.1007/978-3-642-25085-9_27

M3 - Conference contribution

SN - 9783642250842

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

SP - 232

EP - 239

BT - Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 16th Iberoamerican Congress, CIARP 2011, Proceedings

ER -

Peralta B, Soto A. Mixing hierarchical contexts for object recognition. En Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 16th Iberoamerican Congress, CIARP 2011, Proceedings. 2011. p. 232-239. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-25085-9_27