Motion Primitives for Action Recognition

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

304 Downloads (Pure)

Abstract

The number of potential applications has made automatic recognition of human actions a very active research area. Different approaches have been followed based on trajectories through some state space. In this paper we also model an action as a trajectory through a state space, but we represent the actions as a sequence of temporal isolated instances, denoted primitives. These primitives are each defined by four features extracted from motion images. The primitives are recognized in each frame based on a trained classifier resulting in a sequence of primitives. From this sequence we recognize different temporal actions using a probabilistic Edit Distance method. The method is tested on different actions with and without noise and the results show recognition rates of 88.7% and 85.5%, respectively.

Original languageEnglish
Title of host publicationGW 2007 : The 7th International Workshop on Gesture in Human-Computer Interaction and Simulation
EditorsMiguel Sales Dias, Ricardo Jota
Number of pages2
PublisherADETTI/ISCTE
Publication date2007
Pages14-15
ISBN (Print)9789728862053
Publication statusPublished - 2007
EventInternational Workshop on Gesture in Human-Computer Interaction and Simulation - Lisbon, Portugal
Duration: 23 May 200725 May 2007
Conference number: 7

Conference

ConferenceInternational Workshop on Gesture in Human-Computer Interaction and Simulation
Number7
CountryPortugal
CityLisbon
Period23/05/200725/05/2007

Fingerprint

Trajectories
Classifiers

Keywords

  • Gesture recognition
  • Motion primitives

Cite this

Fihl, P., Holte, M. B., & Moeslund, T. B. (2007). Motion Primitives for Action Recognition. In M. Sales Dias, & R. Jota (Eds.), GW 2007: The 7th International Workshop on Gesture in Human-Computer Interaction and Simulation (pp. 14-15). ADETTI/ISCTE.
Fihl, Preben ; Holte, Michael Boelstoft ; Moeslund, Thomas B. / Motion Primitives for Action Recognition. GW 2007: The 7th International Workshop on Gesture in Human-Computer Interaction and Simulation. editor / Miguel Sales Dias ; Ricardo Jota. ADETTI/ISCTE, 2007. pp. 14-15
@inproceedings{b3c69af00dc811dcb676000ea68e967b,
title = "Motion Primitives for Action Recognition",
abstract = "The number of potential applications has made automatic recognition of human actions a very active research area. Different approaches have been followed based on trajectories through some state space. In this paper we also model an action as a trajectory through a state space, but we represent the actions as a sequence of temporal isolated instances, denoted primitives. These primitives are each defined by four features extracted from motion images. The primitives are recognized in each frame based on a trained classifier resulting in a sequence of primitives. From this sequence we recognize different temporal actions using a probabilistic Edit Distance method. The method is tested on different actions with and without noise and the results show recognition rates of 88.7{\%} and 85.5{\%}, respectively.",
keywords = "Gesture recognition, Motion primitives",
author = "Preben Fihl and Holte, {Michael Boelstoft} and Moeslund, {Thomas B.}",
year = "2007",
language = "English",
isbn = "9789728862053",
pages = "14--15",
editor = "{Sales Dias}, Miguel and Ricardo Jota",
booktitle = "GW 2007",
publisher = "ADETTI/ISCTE",

}

Fihl, P, Holte, MB & Moeslund, TB 2007, Motion Primitives for Action Recognition. in M Sales Dias & R Jota (eds), GW 2007: The 7th International Workshop on Gesture in Human-Computer Interaction and Simulation. ADETTI/ISCTE, pp. 14-15, Lisbon, Portugal, 23/05/2007.

Motion Primitives for Action Recognition. / Fihl, Preben; Holte, Michael Boelstoft; Moeslund, Thomas B.

GW 2007: The 7th International Workshop on Gesture in Human-Computer Interaction and Simulation. ed. / Miguel Sales Dias; Ricardo Jota. ADETTI/ISCTE, 2007. p. 14-15.

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Motion Primitives for Action Recognition

AU - Fihl, Preben

AU - Holte, Michael Boelstoft

AU - Moeslund, Thomas B.

PY - 2007

Y1 - 2007

N2 - The number of potential applications has made automatic recognition of human actions a very active research area. Different approaches have been followed based on trajectories through some state space. In this paper we also model an action as a trajectory through a state space, but we represent the actions as a sequence of temporal isolated instances, denoted primitives. These primitives are each defined by four features extracted from motion images. The primitives are recognized in each frame based on a trained classifier resulting in a sequence of primitives. From this sequence we recognize different temporal actions using a probabilistic Edit Distance method. The method is tested on different actions with and without noise and the results show recognition rates of 88.7% and 85.5%, respectively.

AB - The number of potential applications has made automatic recognition of human actions a very active research area. Different approaches have been followed based on trajectories through some state space. In this paper we also model an action as a trajectory through a state space, but we represent the actions as a sequence of temporal isolated instances, denoted primitives. These primitives are each defined by four features extracted from motion images. The primitives are recognized in each frame based on a trained classifier resulting in a sequence of primitives. From this sequence we recognize different temporal actions using a probabilistic Edit Distance method. The method is tested on different actions with and without noise and the results show recognition rates of 88.7% and 85.5%, respectively.

KW - Gesture recognition

KW - Motion primitives

M3 - Article in proceeding

SN - 9789728862053

SP - 14

EP - 15

BT - GW 2007

A2 - Sales Dias, Miguel

A2 - Jota, Ricardo

PB - ADETTI/ISCTE

ER -

Fihl P, Holte MB, Moeslund TB. Motion Primitives for Action Recognition. In Sales Dias M, Jota R, editors, GW 2007: The 7th International Workshop on Gesture in Human-Computer Interaction and Simulation. ADETTI/ISCTE. 2007. p. 14-15