4 Citations (Scopus)

Abstract

In this paper, we consider the problem of fast and efficient indexing techniques for sequences evolving in non-Euclidean spaces. This problem has several applications in the areas of human activity analysis, where there is a need to perform fast search, and recognition in very high dimensional spaces. The problem is made more challenging when representations such as landmarks, contours, and human skeletons etc. are naturally studied in a non-Euclidean setting where even simple operations are much more computationally intensive than their Euclidean counterparts. We propose a geometry and data adaptive symbolic framework that is shown to enable the deployment of fast and accurate algorithms for activity recognition, dynamic texture recognition, motif discovery. Toward this end, we present generalizations of key concepts of piece-wise aggregation and symbolic approximation for the case of non-Euclidean manifolds. We show that one can replace expensive geodesic computations with much faster symbolic computations with little loss of accuracy in activity recognition and discovery applications. The framework is general enough to work across both Euclidean and non-Euclidean spaces, depending on appropriate feature representations without compromising on the ultra-low bandwidth, high speed and high accuracy. The proposed methods are ideally suited for real-time systems and low complexity scenarios.

Original languageEnglish (US)
Pages (from-to)161-173
Number of pages13
JournalInternational Journal of Computer Vision
Volume116
Issue number2
DOIs
StatePublished - Jan 1 2016

Fingerprint

Geometry
Real time systems
Agglomeration
Textures
Bandwidth

Keywords

  • Activity recognition
  • Competitive learning
  • Data mining
  • Differential geometry
  • Manifold sequence indexing

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence
  • Computer Vision and Pattern Recognition

Cite this

Geometry-Based Symbolic Approximation for Fast Sequence Matching on Manifolds. / Anirudh, Rushil; Turaga, Pavan.

In: International Journal of Computer Vision, Vol. 116, No. 2, 01.01.2016, p. 161-173.

Research output: Contribution to journalArticle

@article{90bee0cb3a8748469cda14a289c3397f,
title = "Geometry-Based Symbolic Approximation for Fast Sequence Matching on Manifolds",
abstract = "In this paper, we consider the problem of fast and efficient indexing techniques for sequences evolving in non-Euclidean spaces. This problem has several applications in the areas of human activity analysis, where there is a need to perform fast search, and recognition in very high dimensional spaces. The problem is made more challenging when representations such as landmarks, contours, and human skeletons etc. are naturally studied in a non-Euclidean setting where even simple operations are much more computationally intensive than their Euclidean counterparts. We propose a geometry and data adaptive symbolic framework that is shown to enable the deployment of fast and accurate algorithms for activity recognition, dynamic texture recognition, motif discovery. Toward this end, we present generalizations of key concepts of piece-wise aggregation and symbolic approximation for the case of non-Euclidean manifolds. We show that one can replace expensive geodesic computations with much faster symbolic computations with little loss of accuracy in activity recognition and discovery applications. The framework is general enough to work across both Euclidean and non-Euclidean spaces, depending on appropriate feature representations without compromising on the ultra-low bandwidth, high speed and high accuracy. The proposed methods are ideally suited for real-time systems and low complexity scenarios.",
keywords = "Activity recognition, Competitive learning, Data mining, Differential geometry, Manifold sequence indexing",
author = "Rushil Anirudh and Pavan Turaga",
year = "2016",
month = "1",
day = "1",
doi = "10.1007/s11263-015-0835-8",
language = "English (US)",
volume = "116",
pages = "161--173",
journal = "International Journal of Computer Vision",
issn = "0920-5691",
publisher = "Springer Netherlands",
number = "2",

}

TY - JOUR

T1 - Geometry-Based Symbolic Approximation for Fast Sequence Matching on Manifolds

AU - Anirudh, Rushil

AU - Turaga, Pavan

PY - 2016/1/1

Y1 - 2016/1/1

N2 - In this paper, we consider the problem of fast and efficient indexing techniques for sequences evolving in non-Euclidean spaces. This problem has several applications in the areas of human activity analysis, where there is a need to perform fast search, and recognition in very high dimensional spaces. The problem is made more challenging when representations such as landmarks, contours, and human skeletons etc. are naturally studied in a non-Euclidean setting where even simple operations are much more computationally intensive than their Euclidean counterparts. We propose a geometry and data adaptive symbolic framework that is shown to enable the deployment of fast and accurate algorithms for activity recognition, dynamic texture recognition, motif discovery. Toward this end, we present generalizations of key concepts of piece-wise aggregation and symbolic approximation for the case of non-Euclidean manifolds. We show that one can replace expensive geodesic computations with much faster symbolic computations with little loss of accuracy in activity recognition and discovery applications. The framework is general enough to work across both Euclidean and non-Euclidean spaces, depending on appropriate feature representations without compromising on the ultra-low bandwidth, high speed and high accuracy. The proposed methods are ideally suited for real-time systems and low complexity scenarios.

AB - In this paper, we consider the problem of fast and efficient indexing techniques for sequences evolving in non-Euclidean spaces. This problem has several applications in the areas of human activity analysis, where there is a need to perform fast search, and recognition in very high dimensional spaces. The problem is made more challenging when representations such as landmarks, contours, and human skeletons etc. are naturally studied in a non-Euclidean setting where even simple operations are much more computationally intensive than their Euclidean counterparts. We propose a geometry and data adaptive symbolic framework that is shown to enable the deployment of fast and accurate algorithms for activity recognition, dynamic texture recognition, motif discovery. Toward this end, we present generalizations of key concepts of piece-wise aggregation and symbolic approximation for the case of non-Euclidean manifolds. We show that one can replace expensive geodesic computations with much faster symbolic computations with little loss of accuracy in activity recognition and discovery applications. The framework is general enough to work across both Euclidean and non-Euclidean spaces, depending on appropriate feature representations without compromising on the ultra-low bandwidth, high speed and high accuracy. The proposed methods are ideally suited for real-time systems and low complexity scenarios.

KW - Activity recognition

KW - Competitive learning

KW - Data mining

KW - Differential geometry

KW - Manifold sequence indexing

UR - http://www.scopus.com/inward/record.url?scp=84955216817&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84955216817&partnerID=8YFLogxK

U2 - 10.1007/s11263-015-0835-8

DO - 10.1007/s11263-015-0835-8

M3 - Article

VL - 116

SP - 161

EP - 173

JO - International Journal of Computer Vision

JF - International Journal of Computer Vision

SN - 0920-5691

IS - 2

ER -