Human action classification using SVM_2K classifier on motion features

Hongying Meng; Nick Pears; Chris Bailey

Human action classification using SVM_2K classifier on motion features

Computer Science

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

Abstract

In this paper, we study the human action classification problem based on motion features directly extracted from video. In order to implement a fast classification system, we select simple features that can be obtained from non-intensive computation. We also introduce the new SVM_2K classifier that can achieve improved performance over a standard SVM by combining two types of motion feature vector together. After learning, classification can be implemented very quickly because SVM_2K is a linear classifier. Experimental results demonstrate the method to be efficient and may be used in real-time human action classification systems.

Original language	English
Title of host publication	MULTIMEDIA CONTENT REPRESENTATION, CLASSIFICATION AND SECURITY
Editors	B Gunsel, AK Jain, AM Tekalp, B Sankur
Place of Publication	BERLIN
Publisher	Springer
Pages	458-465
Number of pages	8
ISBN (Print)	3-540-39392-7
Publication status	Published - 2006
Event	International Workshop on Multimedia Content Representation, Classification and Security (MRCS 2006) - Istanbul Duration: 11 Sept 2006 → 13 Sept 2006

Conference

Conference	International Workshop on Multimedia Content Representation, Classification and Security (MRCS 2006)
City	Istanbul
Period	11/09/06 → 13/09/06

Keywords

RECOGNITION

Cite this

@inproceedings{eb96ddb4bc7b4686975659cd3c353ed6,

title = "Human action classification using SVM_2K classifier on motion features",

abstract = "In this paper, we study the human action classification problem based on motion features directly extracted from video. In order to implement a fast classification system, we select simple features that can be obtained from non-intensive computation. We also introduce the new SVM_2K classifier that can achieve improved performance over a standard SVM by combining two types of motion feature vector together. After learning, classification can be implemented very quickly because SVM_2K is a linear classifier. Experimental results demonstrate the method to be efficient and may be used in real-time human action classification systems.",

keywords = "RECOGNITION",

author = "Hongying Meng and Nick Pears and Chris Bailey",

year = "2006",

language = "English",

isbn = "3-540-39392-7",

pages = "458--465",

editor = "B Gunsel and AK Jain and AM Tekalp and B Sankur",

booktitle = "MULTIMEDIA CONTENT REPRESENTATION, CLASSIFICATION AND SECURITY",

publisher = "Springer",

address = "Germany",

note = "International Workshop on Multimedia Content Representation, Classification and Security (MRCS 2006) ; Conference date: 11-09-2006 Through 13-09-2006",

}

Meng, H, Pears, N & Bailey, C 2006, Human action classification using SVM_2K classifier on motion features. in B Gunsel, AK Jain, AM Tekalp & B Sankur (eds), MULTIMEDIA CONTENT REPRESENTATION, CLASSIFICATION AND SECURITY. Springer, BERLIN, pp. 458-465, International Workshop on Multimedia Content Representation, Classification and Security (MRCS 2006), Istanbul, 11/09/06.

TY - GEN

T1 - Human action classification using SVM_2K classifier on motion features

AU - Meng, Hongying

AU - Pears, Nick

AU - Bailey, Chris

PY - 2006

Y1 - 2006

N2 - In this paper, we study the human action classification problem based on motion features directly extracted from video. In order to implement a fast classification system, we select simple features that can be obtained from non-intensive computation. We also introduce the new SVM_2K classifier that can achieve improved performance over a standard SVM by combining two types of motion feature vector together. After learning, classification can be implemented very quickly because SVM_2K is a linear classifier. Experimental results demonstrate the method to be efficient and may be used in real-time human action classification systems.

AB - In this paper, we study the human action classification problem based on motion features directly extracted from video. In order to implement a fast classification system, we select simple features that can be obtained from non-intensive computation. We also introduce the new SVM_2K classifier that can achieve improved performance over a standard SVM by combining two types of motion feature vector together. After learning, classification can be implemented very quickly because SVM_2K is a linear classifier. Experimental results demonstrate the method to be efficient and may be used in real-time human action classification systems.

KW - RECOGNITION

M3 - Conference contribution

SN - 3-540-39392-7

SP - 458

EP - 465

BT - MULTIMEDIA CONTENT REPRESENTATION, CLASSIFICATION AND SECURITY

A2 - Gunsel, B

A2 - Jain, AK

A2 - Tekalp, AM

A2 - Sankur, B

PB - Springer

CY - BERLIN

T2 - International Workshop on Multimedia Content Representation, Classification and Security (MRCS 2006)

Y2 - 11 September 2006 through 13 September 2006

ER -