Generalized Dictionaries for Multiple Instance Learning

Ashish Shrivastava; Jaishanker K. Pillai; Rama Chellappa

doi:10.1007/s11263-015-0831-z

Generalized Dictionaries for Multiple Instance Learning

Ashish Shrivastava, Jaishanker K. Pillai, Rama Chellappa

School of Engineering, Electrical & Computer Engineering

Rutgers, The State University

Research output: Contribution to journal › Article › peer-review

22 Scopus citations

Abstract

We present a multi-class multiple instance learning (MIL) algorithm using the dictionary learning framework where the data is given in the form of bags. Each bag contains multiple samples, called instances, out of which at least one belongs to the class of the bag. We propose a noisy-OR model and a generalized mean-based optimization framework for learning the dictionaries in the feature space. The proposed method can be viewed as a generalized dictionary learning algorithm since it reduces to a novel discriminative dictionary learning framework when there is only one instance in each bag. Various experiments using popular vision-related MIL datasets as well as the UNBC-McMaster Pain Shoulder Archive database show that the proposed method performs significantly better than the existing methods.

Original language	American English
Pages (from-to)	288-305
Number of pages	18
Journal	International Journal of Computer Vision
Volume	114
Issue number	2-3
DOIs	https://doi.org/10.1007/s11263-015-0831-z
State	Published - Sep 22 2015

ASJC Scopus subject areas

Software
Computer Vision and Pattern Recognition
Artificial Intelligence

Keywords

Dictionary learning
Multiple instance learning
Object recognition
Pain detection
Sparse coding

Access to Document

10.1007/s11263-015-0831-z

Cite this

@article{096ffdadb19b42ccafc76202776ffe7d,

title = "Generalized Dictionaries for Multiple Instance Learning",

abstract = "We present a multi-class multiple instance learning (MIL) algorithm using the dictionary learning framework where the data is given in the form of bags. Each bag contains multiple samples, called instances, out of which at least one belongs to the class of the bag. We propose a noisy-OR model and a generalized mean-based optimization framework for learning the dictionaries in the feature space. The proposed method can be viewed as a generalized dictionary learning algorithm since it reduces to a novel discriminative dictionary learning framework when there is only one instance in each bag. Various experiments using popular vision-related MIL datasets as well as the UNBC-McMaster Pain Shoulder Archive database show that the proposed method performs significantly better than the existing methods.",

keywords = "Dictionary learning, Multiple instance learning, Object recognition, Pain detection, Sparse coding",

author = "Ashish Shrivastava and Pillai, \{Jaishanker K.\} and Rama Chellappa",

note = "Publisher Copyright: {\textcopyright} 2015, Springer Science+Business Media New York.",

year = "2015",

month = sep,

day = "22",

doi = "10.1007/s11263-015-0831-z",

language = "American English",

volume = "114",

pages = "288--305",

journal = "International Journal of Computer Vision",

issn = "0920-5691",

publisher = "Springer Netherlands",

number = "2-3",

}

TY - JOUR

T1 - Generalized Dictionaries for Multiple Instance Learning

AU - Shrivastava, Ashish

AU - Pillai, Jaishanker K.

AU - Chellappa, Rama

PY - 2015/9/22

Y1 - 2015/9/22

N2 - We present a multi-class multiple instance learning (MIL) algorithm using the dictionary learning framework where the data is given in the form of bags. Each bag contains multiple samples, called instances, out of which at least one belongs to the class of the bag. We propose a noisy-OR model and a generalized mean-based optimization framework for learning the dictionaries in the feature space. The proposed method can be viewed as a generalized dictionary learning algorithm since it reduces to a novel discriminative dictionary learning framework when there is only one instance in each bag. Various experiments using popular vision-related MIL datasets as well as the UNBC-McMaster Pain Shoulder Archive database show that the proposed method performs significantly better than the existing methods.

AB - We present a multi-class multiple instance learning (MIL) algorithm using the dictionary learning framework where the data is given in the form of bags. Each bag contains multiple samples, called instances, out of which at least one belongs to the class of the bag. We propose a noisy-OR model and a generalized mean-based optimization framework for learning the dictionaries in the feature space. The proposed method can be viewed as a generalized dictionary learning algorithm since it reduces to a novel discriminative dictionary learning framework when there is only one instance in each bag. Various experiments using popular vision-related MIL datasets as well as the UNBC-McMaster Pain Shoulder Archive database show that the proposed method performs significantly better than the existing methods.

KW - Dictionary learning

KW - Multiple instance learning

KW - Object recognition

KW - Pain detection

KW - Sparse coding

UR - http://www.scopus.com/inward/record.url?scp=84939564619&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84939564619&partnerID=8YFLogxK

U2 - 10.1007/s11263-015-0831-z

DO - 10.1007/s11263-015-0831-z

M3 - Article

SN - 0920-5691

VL - 114

SP - 288

EP - 305

JO - International Journal of Computer Vision

JF - International Journal of Computer Vision

IS - 2-3

ER -

Generalized Dictionaries for Multiple Instance Learning

Abstract

ASJC Scopus subject areas

Keywords

Access to Document

Other files and links

Fingerprint

Cite this