Dynamic template tracking and recognition

Rizwan Chaudhry, Gregory Hager, René Vidal

Research output: Contribution to journalArticle

Abstract

In this paper we address the problem of tracking non-rigid objects whose local appearance and motion changes as a function of time. This class of objects includes dynamic textures such as steam, fire, smoke, water, etc., as well as articulated objects such as humans performing various actions. We model the temporal evolution of the object's appearance/motion using a linear dynamical system. We learn such models from sample videos and use them as dynamic templates for tracking objects in novel videos. We pose the problem of tracking a dynamic non-rigid object in the current frame as a maximum a-posteriori estimate of the location of the object and the latent state of the dynamical system, given the current image features and the best estimate of the state in the previous frame. The advantage of our approach is that we can specify a-priori the type of texture to be tracked in the scene by using previously trained models for the dynamics of these textures. Our framework naturally generalizes common tracking methods such as SSD and kernelbased tracking from static templates to dynamic templates. We test our algorithm on synthetic as well as real examples of dynamic textures and show that our simple dynamics-based trackers perform at par if not better than the state-of-the-art. Since our approach is general and applicable to any image feature, we also apply it to the problem of human action tracking and build action-specific optical flow trackers that per-form better than the state-of-the-art when tracking a human performing a particular action. Finally, since our approach is generative, we can use a-priori trained trackers for different texture or action classes to simultaneously track and recognize the texture or action in the video.

Original languageEnglish (US)
Pages (from-to)19-48
Number of pages30
JournalInternational Journal of Computer Vision
Volume105
Issue number1
DOIs
StatePublished - Oct 2013

Fingerprint

Textures
Dynamical systems
Optical flows
Smoke
Fires
Steam
Water

Keywords

  • Dynamic templates
  • Dynamic textures
  • Human actions
  • Linear dynamical systems
  • Recognition
  • Tracking

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence
  • Computer Vision and Pattern Recognition

Cite this

Dynamic template tracking and recognition. / Chaudhry, Rizwan; Hager, Gregory; Vidal, René.

In: International Journal of Computer Vision, Vol. 105, No. 1, 10.2013, p. 19-48.

Research output: Contribution to journalArticle

Chaudhry, Rizwan ; Hager, Gregory ; Vidal, René. / Dynamic template tracking and recognition. In: International Journal of Computer Vision. 2013 ; Vol. 105, No. 1. pp. 19-48.
@article{bfc03b2119234a7aa2a4fbc1da3d68dd,
title = "Dynamic template tracking and recognition",
abstract = "In this paper we address the problem of tracking non-rigid objects whose local appearance and motion changes as a function of time. This class of objects includes dynamic textures such as steam, fire, smoke, water, etc., as well as articulated objects such as humans performing various actions. We model the temporal evolution of the object's appearance/motion using a linear dynamical system. We learn such models from sample videos and use them as dynamic templates for tracking objects in novel videos. We pose the problem of tracking a dynamic non-rigid object in the current frame as a maximum a-posteriori estimate of the location of the object and the latent state of the dynamical system, given the current image features and the best estimate of the state in the previous frame. The advantage of our approach is that we can specify a-priori the type of texture to be tracked in the scene by using previously trained models for the dynamics of these textures. Our framework naturally generalizes common tracking methods such as SSD and kernelbased tracking from static templates to dynamic templates. We test our algorithm on synthetic as well as real examples of dynamic textures and show that our simple dynamics-based trackers perform at par if not better than the state-of-the-art. Since our approach is general and applicable to any image feature, we also apply it to the problem of human action tracking and build action-specific optical flow trackers that per-form better than the state-of-the-art when tracking a human performing a particular action. Finally, since our approach is generative, we can use a-priori trained trackers for different texture or action classes to simultaneously track and recognize the texture or action in the video.",
keywords = "Dynamic templates, Dynamic textures, Human actions, Linear dynamical systems, Recognition, Tracking",
author = "Rizwan Chaudhry and Gregory Hager and Ren{\'e} Vidal",
year = "2013",
month = "10",
doi = "10.1007/s11263-013-0625-0",
language = "English (US)",
volume = "105",
pages = "19--48",
journal = "International Journal of Computer Vision",
issn = "0920-5691",
publisher = "Springer Netherlands",
number = "1",

}

TY - JOUR

T1 - Dynamic template tracking and recognition

AU - Chaudhry, Rizwan

AU - Hager, Gregory

AU - Vidal, René

PY - 2013/10

Y1 - 2013/10

N2 - In this paper we address the problem of tracking non-rigid objects whose local appearance and motion changes as a function of time. This class of objects includes dynamic textures such as steam, fire, smoke, water, etc., as well as articulated objects such as humans performing various actions. We model the temporal evolution of the object's appearance/motion using a linear dynamical system. We learn such models from sample videos and use them as dynamic templates for tracking objects in novel videos. We pose the problem of tracking a dynamic non-rigid object in the current frame as a maximum a-posteriori estimate of the location of the object and the latent state of the dynamical system, given the current image features and the best estimate of the state in the previous frame. The advantage of our approach is that we can specify a-priori the type of texture to be tracked in the scene by using previously trained models for the dynamics of these textures. Our framework naturally generalizes common tracking methods such as SSD and kernelbased tracking from static templates to dynamic templates. We test our algorithm on synthetic as well as real examples of dynamic textures and show that our simple dynamics-based trackers perform at par if not better than the state-of-the-art. Since our approach is general and applicable to any image feature, we also apply it to the problem of human action tracking and build action-specific optical flow trackers that per-form better than the state-of-the-art when tracking a human performing a particular action. Finally, since our approach is generative, we can use a-priori trained trackers for different texture or action classes to simultaneously track and recognize the texture or action in the video.

AB - In this paper we address the problem of tracking non-rigid objects whose local appearance and motion changes as a function of time. This class of objects includes dynamic textures such as steam, fire, smoke, water, etc., as well as articulated objects such as humans performing various actions. We model the temporal evolution of the object's appearance/motion using a linear dynamical system. We learn such models from sample videos and use them as dynamic templates for tracking objects in novel videos. We pose the problem of tracking a dynamic non-rigid object in the current frame as a maximum a-posteriori estimate of the location of the object and the latent state of the dynamical system, given the current image features and the best estimate of the state in the previous frame. The advantage of our approach is that we can specify a-priori the type of texture to be tracked in the scene by using previously trained models for the dynamics of these textures. Our framework naturally generalizes common tracking methods such as SSD and kernelbased tracking from static templates to dynamic templates. We test our algorithm on synthetic as well as real examples of dynamic textures and show that our simple dynamics-based trackers perform at par if not better than the state-of-the-art. Since our approach is general and applicable to any image feature, we also apply it to the problem of human action tracking and build action-specific optical flow trackers that per-form better than the state-of-the-art when tracking a human performing a particular action. Finally, since our approach is generative, we can use a-priori trained trackers for different texture or action classes to simultaneously track and recognize the texture or action in the video.

KW - Dynamic templates

KW - Dynamic textures

KW - Human actions

KW - Linear dynamical systems

KW - Recognition

KW - Tracking

UR - http://www.scopus.com/inward/record.url?scp=84892882025&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84892882025&partnerID=8YFLogxK

U2 - 10.1007/s11263-013-0625-0

DO - 10.1007/s11263-013-0625-0

M3 - Article

AN - SCOPUS:84892882025

VL - 105

SP - 19

EP - 48

JO - International Journal of Computer Vision

JF - International Journal of Computer Vision

SN - 0920-5691

IS - 1

ER -