Pornographic images recognition based on spatial pyramid partition and multi-instance ensemble learning

Daxiang Li, Na Li, Jing Wang, Tingge Zhu

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

For tackling the problem of pornographic image recognition, a novel multi-instance learning (MIL) algorithm is proposed by using extreme learning machine (ELM) and classifiers ensemble. Firstly, a spatial pyramid partition-based (SPP) multi-instance modeling technique has been deployed to transform the pornographic images recognition problem into a typical MIL problem. The method has deployed a bag corresponding to an image and an instance corresponding to each partitioned sub-block described by low-level visual features (i.e. color, texture and shape). Secondly, a collection of visual word (VW) has been generated by using hierarchical k-mean clustering method, and then based on the fuzzy membership function between instance and VW, a fuzzy histogram fusion-based metadata calculation method has been proposed to convert each bag to a single sample, which allows the MIL problem to be solved directly by a standard single instance learning (SIL) machine. Finally, by using ELM, a group of base classifiers with different number of hidden nodes have been constructed, and their weights bas been dynamically determined by using performance weighting rule. Therefore, the strategy of classifiers ensemble is used to improve the overall adaptability of proposed ELMCE-MIL algorithm. Experimental results have shown that the method is robust, and its performance is superior to other similar algorithms.
LanguageEnglish
Pages214-223
Number of pages10
JournalKnowledge-Based Systems
Volume84
Early online date28 Apr 2015
DOIs
Publication statusPublished - Aug 2015

Fingerprint

Image recognition
Learning systems
Classifiers
Learning algorithms
Membership functions
Metadata
Fusion reactions
Textures
Color
Machine learning
Classifier
Ensemble learning
Learning algorithm

Cite this

Li, Daxiang ; Li, Na ; Wang, Jing ; Zhu, Tingge. / Pornographic images recognition based on spatial pyramid partition and multi-instance ensemble learning. In: Knowledge-Based Systems. 2015 ; Vol. 84. pp. 214-223.
@article{8a046b003a5c4ea3822665fae69653af,
title = "Pornographic images recognition based on spatial pyramid partition and multi-instance ensemble learning",
abstract = "For tackling the problem of pornographic image recognition, a novel multi-instance learning (MIL) algorithm is proposed by using extreme learning machine (ELM) and classifiers ensemble. Firstly, a spatial pyramid partition-based (SPP) multi-instance modeling technique has been deployed to transform the pornographic images recognition problem into a typical MIL problem. The method has deployed a bag corresponding to an image and an instance corresponding to each partitioned sub-block described by low-level visual features (i.e. color, texture and shape). Secondly, a collection of visual word (VW) has been generated by using hierarchical k-mean clustering method, and then based on the fuzzy membership function between instance and VW, a fuzzy histogram fusion-based metadata calculation method has been proposed to convert each bag to a single sample, which allows the MIL problem to be solved directly by a standard single instance learning (SIL) machine. Finally, by using ELM, a group of base classifiers with different number of hidden nodes have been constructed, and their weights bas been dynamically determined by using performance weighting rule. Therefore, the strategy of classifiers ensemble is used to improve the overall adaptability of proposed ELMCE-MIL algorithm. Experimental results have shown that the method is robust, and its performance is superior to other similar algorithms.",
keywords = "Multi-instance learning (MIL), pornographic images recognition, extreme learning machine",
author = "Daxiang Li and Na Li and Jing Wang and Tingge Zhu",
year = "2015",
month = "8",
doi = "10.1016/j.knosys.2015.04.014",
language = "English",
volume = "84",
pages = "214--223",
journal = "Knowledge-Based Systems",
issn = "0950-7051",
publisher = "Elsevier",

}

Pornographic images recognition based on spatial pyramid partition and multi-instance ensemble learning. / Li, Daxiang; Li, Na; Wang, Jing; Zhu, Tingge.

In: Knowledge-Based Systems, Vol. 84, 08.2015, p. 214-223.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Pornographic images recognition based on spatial pyramid partition and multi-instance ensemble learning

AU - Li, Daxiang

AU - Li, Na

AU - Wang, Jing

AU - Zhu, Tingge

PY - 2015/8

Y1 - 2015/8

N2 - For tackling the problem of pornographic image recognition, a novel multi-instance learning (MIL) algorithm is proposed by using extreme learning machine (ELM) and classifiers ensemble. Firstly, a spatial pyramid partition-based (SPP) multi-instance modeling technique has been deployed to transform the pornographic images recognition problem into a typical MIL problem. The method has deployed a bag corresponding to an image and an instance corresponding to each partitioned sub-block described by low-level visual features (i.e. color, texture and shape). Secondly, a collection of visual word (VW) has been generated by using hierarchical k-mean clustering method, and then based on the fuzzy membership function between instance and VW, a fuzzy histogram fusion-based metadata calculation method has been proposed to convert each bag to a single sample, which allows the MIL problem to be solved directly by a standard single instance learning (SIL) machine. Finally, by using ELM, a group of base classifiers with different number of hidden nodes have been constructed, and their weights bas been dynamically determined by using performance weighting rule. Therefore, the strategy of classifiers ensemble is used to improve the overall adaptability of proposed ELMCE-MIL algorithm. Experimental results have shown that the method is robust, and its performance is superior to other similar algorithms.

AB - For tackling the problem of pornographic image recognition, a novel multi-instance learning (MIL) algorithm is proposed by using extreme learning machine (ELM) and classifiers ensemble. Firstly, a spatial pyramid partition-based (SPP) multi-instance modeling technique has been deployed to transform the pornographic images recognition problem into a typical MIL problem. The method has deployed a bag corresponding to an image and an instance corresponding to each partitioned sub-block described by low-level visual features (i.e. color, texture and shape). Secondly, a collection of visual word (VW) has been generated by using hierarchical k-mean clustering method, and then based on the fuzzy membership function between instance and VW, a fuzzy histogram fusion-based metadata calculation method has been proposed to convert each bag to a single sample, which allows the MIL problem to be solved directly by a standard single instance learning (SIL) machine. Finally, by using ELM, a group of base classifiers with different number of hidden nodes have been constructed, and their weights bas been dynamically determined by using performance weighting rule. Therefore, the strategy of classifiers ensemble is used to improve the overall adaptability of proposed ELMCE-MIL algorithm. Experimental results have shown that the method is robust, and its performance is superior to other similar algorithms.

KW - Multi-instance learning (MIL)

KW - pornographic images recognition

KW - extreme learning machine

U2 - 10.1016/j.knosys.2015.04.014

DO - 10.1016/j.knosys.2015.04.014

M3 - Article

VL - 84

SP - 214

EP - 223

JO - Knowledge-Based Systems

T2 - Knowledge-Based Systems

JF - Knowledge-Based Systems

SN - 0950-7051

ER -