GPU Hierarchical Quilted Self Organizing Maps for multimedia understanding

Youssef S G Nashed

doi:10.1109/ISM.2012.102

GPU Hierarchical Quilted Self Organizing Maps for multimedia understanding

Youssef S G Nashed^*

^*Corresponding author for this work

Electrical and Computer Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

1 Scopus citations

Abstract

It is well established that the human brain outperforms current computers, concerning pattern recognition tasks, through the collaborative processing of simple building units (neurons). In this work we expand an abstracted model of the neocortex called Hierarchical Quilted Self Organizing Map, benefiting from the parallel power of current Graphical Processing Units, to achieve realtime understanding and classification of spatio-temporal sensory information. We also propose an improvement on the original model that allows the learning rate to be automatically adapted according to the input training data available. The overall system is tested on the task of gesture recognition from a Microsoft Kinect publicly available dataset.

Original language	English (US)
Title of host publication	Proceedings - 2012 IEEE International Symposium on Multimedia, ISM 2012
Pages	491-492
Number of pages	2
DOIs	https://doi.org/10.1109/ISM.2012.102
State	Published - Dec 1 2012
Event	14th IEEE International Symposium on Multimedia, ISM 2012 - Irvine, CA, United States Duration: Dec 10 2012 → Dec 12 2012

Other

Other	14th IEEE International Symposium on Multimedia, ISM 2012
Country/Territory	United States
City	Irvine, CA
Period	12/10/12 → 12/12/12

Keywords

GPU
Memory prediction framework
Neural networks
Self Organizing Maps
Temporal classification

ASJC Scopus subject areas

Computer Graphics and Computer-Aided Design
Computer Vision and Pattern Recognition
Software

Access to Document

10.1109/ISM.2012.102

Cite this

@inproceedings{ff9ba342d9414709b184d2b5c18e7446,

title = "GPU Hierarchical Quilted Self Organizing Maps for multimedia understanding",

abstract = "It is well established that the human brain outperforms current computers, concerning pattern recognition tasks, through the collaborative processing of simple building units (neurons). In this work we expand an abstracted model of the neocortex called Hierarchical Quilted Self Organizing Map, benefiting from the parallel power of current Graphical Processing Units, to achieve realtime understanding and classification of spatio-temporal sensory information. We also propose an improvement on the original model that allows the learning rate to be automatically adapted according to the input training data available. The overall system is tested on the task of gesture recognition from a Microsoft Kinect publicly available dataset.",

keywords = "GPU, Memory prediction framework, Neural networks, Self Organizing Maps, Temporal classification",

author = "Nashed, {Youssef S G}",

year = "2012",

month = dec,

day = "1",

doi = "10.1109/ISM.2012.102",

language = "English (US)",

isbn = "9780769548753",

pages = "491--492",

booktitle = "Proceedings - 2012 IEEE International Symposium on Multimedia, ISM 2012",

note = "14th IEEE International Symposium on Multimedia, ISM 2012 ; Conference date: 10-12-2012 Through 12-12-2012",

}

TY - GEN

T1 - GPU Hierarchical Quilted Self Organizing Maps for multimedia understanding

AU - Nashed, Youssef S G

PY - 2012/12/1

Y1 - 2012/12/1

N2 - It is well established that the human brain outperforms current computers, concerning pattern recognition tasks, through the collaborative processing of simple building units (neurons). In this work we expand an abstracted model of the neocortex called Hierarchical Quilted Self Organizing Map, benefiting from the parallel power of current Graphical Processing Units, to achieve realtime understanding and classification of spatio-temporal sensory information. We also propose an improvement on the original model that allows the learning rate to be automatically adapted according to the input training data available. The overall system is tested on the task of gesture recognition from a Microsoft Kinect publicly available dataset.

AB - It is well established that the human brain outperforms current computers, concerning pattern recognition tasks, through the collaborative processing of simple building units (neurons). In this work we expand an abstracted model of the neocortex called Hierarchical Quilted Self Organizing Map, benefiting from the parallel power of current Graphical Processing Units, to achieve realtime understanding and classification of spatio-temporal sensory information. We also propose an improvement on the original model that allows the learning rate to be automatically adapted according to the input training data available. The overall system is tested on the task of gesture recognition from a Microsoft Kinect publicly available dataset.

KW - GPU

KW - Memory prediction framework

KW - Neural networks

KW - Self Organizing Maps

KW - Temporal classification

UR - http://www.scopus.com/inward/record.url?scp=84874274857&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84874274857&partnerID=8YFLogxK

U2 - 10.1109/ISM.2012.102

DO - 10.1109/ISM.2012.102

M3 - Conference contribution

AN - SCOPUS:84874274857

SN - 9780769548753

SP - 491

EP - 492

BT - Proceedings - 2012 IEEE International Symposium on Multimedia, ISM 2012

T2 - 14th IEEE International Symposium on Multimedia, ISM 2012

Y2 - 10 December 2012 through 12 December 2012

ER -

GPU Hierarchical Quilted Self Organizing Maps for multimedia understanding

Abstract

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this