Hierarchical Self-Organizing Maps System for Action Classification

Zahra Gharaee; Peter Gärdenfors; Magnus Johnsson

doi:10.5220/0006199305830590

Hierarchical Self-Organizing Maps System for Action Classification

Zahra Gharaee, Peter Gärdenfors, Magnus Johnsson

Research output: Chapter in Book/Report/Conference proceeding › Paper in conference proceeding › peer-review

Abstract

We present a novel action recognition system that is able to learn how to recognize and classify actions. Our system employs a three-layered hierarchy of Self-Organizing Maps together with a supervised neural network for labelling the actions. We have evaluated our system in an experiments consisting of ten different actions from a publicly available data set. The results are encouraging with 83% correctly classified actions based on the actor’s spatial trajectory.

Original language	English
Title of host publication	Proceedings of the 9th International Conference on Agents and Artiﬁcial Intelligence (ICAART 2017)
Publisher	SciTePress
Pages	583-590
Number of pages	8
ISBN (Electronic)	978-989-758-220-2
DOIs	https://doi.org/10.5220/0006199305830590
Publication status	Published - 2017
Event	ICAART 2017-International Conference on Agents and Artificial Intelligence - Holiday Inn Porto Gaia, Porto, Portugal Duration: 2017 Feb 24 → 2017 Feb 26

Conference

Conference	ICAART 2017-International Conference on Agents and Artificial Intelligence
Country/Territory	Portugal
City	Porto
Period	2017/02/24 → 2017/02/26

Subject classification (UKÄ)

Robotics and automation
Computer Systems

Access to Document

10.5220/0006199305830590

1 Doctoral Thesis (compilation)

Action in Mind: A Neural Network Approach to Action Recognition and Segmentation
Gharaee, Z., 2018 Mar 15, Lund: Lund University Cognitive Science. 166 p.
Research output: Thesis › Doctoral Thesis (compilation)

Open Access
File

2 Finished

CCL: Thinking in Time: Cognition, Communication and Learning
Sahlén, B. (PI), Gärdenfors, P. (PI), Hesslow, G. (Researcher), Lindgren, M. (PI), Strömqvist, S. (PI), Byström, I. (Administrator), Sjöstrand, E. (Administrator), Mårtensson, J. (Researcher), Andersson, A. (Researcher), Hansson, K. (Researcher), Gullberg, M. (Researcher), Johansson, R. (Researcher), Johansson, M. (Researcher), Falck, A. (Researcher), Roll, M. (Researcher), Ståhlberg, F. (Researcher), Horne, M. (Researcher), Blomberg, F. (Researcher), Johansson, V. (Researcher), Nirme, J. (Research student), Gulz, A. (Researcher), van de Weijer, J. (Researcher), Paradis, C. (Researcher), Tärning, B. (Research student), Bramao, I. (Researcher), Sayehli, S. (Researcher), Löhndorf, S. (Research student), Willners, C. (Researcher), Blomberg, J. (Research assistant), Schötz, S. (Researcher), Haake, M. (Researcher), Brännström, J. (Researcher), Grenner, E. (Research student), Holsanova, J. (Researcher), Wengelin, Å. (Researcher), Johnsson, M. (Researcher), Winberg, S. (Researcher), Balkenius, C. (Researcher), Gharaee, Z. (Researcher) & Bååth, R. (Research student)
Swedish Research Council
2008/01/01 → 2018/12/31
Project: Research
Ikaros: An infrastructure for system level modelling of the brain
Johansson, B. (Researcher), Balkenius, C. (PI), Johnsson, M. (Researcher), Bååth, R. (Researcher), Winberg, S. (Researcher), Gharaee, Z. (Researcher), Tjøstheim, T. A. (Research student) & Krantz, A. (Researcher)
2001/01/01 → 2023/12/31
Project: Research

Cite this

@inproceedings{98c8370a81704acb918e37cb31933e8b,

title = "Hierarchical Self-Organizing Maps System for Action Classification",

abstract = "We present a novel action recognition system that is able to learn how to recognize and classify actions. Our system employs a three-layered hierarchy of Self-Organizing Maps together with a supervised neural network for labelling the actions. We have evaluated our system in an experiments consisting of ten different actions from a publicly available data set. The results are encouraging with 83% correctly classified actions based on the actor{\textquoteright}s spatial trajectory.",

author = "Zahra Gharaee and Peter G{\"a}rdenfors and Magnus Johnsson",

year = "2017",

doi = "10.5220/0006199305830590",

language = "English",

pages = "583--590",

booktitle = "Proceedings of the 9th International Conference on Agents and Artiﬁcial Intelligence (ICAART 2017)",

publisher = "SciTePress",

note = "ICAART 2017-International Conference on Agents and Artificial Intelligence ; Conference date: 24-02-2017 Through 26-02-2017",

}

Gharaee, Z, Gärdenfors, P & Johnsson, M 2017, Hierarchical Self-Organizing Maps System for Action Classification. in Proceedings of the 9th International Conference on Agents and Artiﬁcial Intelligence (ICAART 2017). SciTePress, pp. 583-590, ICAART 2017-International Conference on Agents and Artificial Intelligence, Porto, Portugal, 2017/02/24. https://doi.org/10.5220/0006199305830590

TY - GEN

T1 - Hierarchical Self-Organizing Maps System for Action Classification

AU - Gharaee, Zahra

AU - Gärdenfors, Peter

AU - Johnsson, Magnus

PY - 2017

Y1 - 2017

N2 - We present a novel action recognition system that is able to learn how to recognize and classify actions. Our system employs a three-layered hierarchy of Self-Organizing Maps together with a supervised neural network for labelling the actions. We have evaluated our system in an experiments consisting of ten different actions from a publicly available data set. The results are encouraging with 83% correctly classified actions based on the actor’s spatial trajectory.

AB - We present a novel action recognition system that is able to learn how to recognize and classify actions. Our system employs a three-layered hierarchy of Self-Organizing Maps together with a supervised neural network for labelling the actions. We have evaluated our system in an experiments consisting of ten different actions from a publicly available data set. The results are encouraging with 83% correctly classified actions based on the actor’s spatial trajectory.

U2 - 10.5220/0006199305830590

DO - 10.5220/0006199305830590

M3 - Paper in conference proceeding

SP - 583

EP - 590

BT - Proceedings of the 9th International Conference on Agents and Artiﬁcial Intelligence (ICAART 2017)

PB - SciTePress

T2 - ICAART 2017-International Conference on Agents and Artificial Intelligence

Y2 - 24 February 2017 through 26 February 2017

ER -

Hierarchical Self-Organizing Maps System for Action Classification

Abstract

Conference

Subject classification (UKÄ)

Access to Document

Fingerprint

Research output

Action in Mind: A Neural Network Approach to Action Recognition and Segmentation

Projects

CCL: Thinking in Time: Cognition, Communication and Learning

Ikaros: An infrastructure for system level modelling of the brain

Cite this