Recognition by Enhanced Bag of Words Model via Topographic ICA

Min Jing, Hui Wang, Kathy Clawson, SA Coleman, Shuwei Chen, Jun Liu, Bryan Scotney

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The Bag-of-Words (BoW) model has been increasingly applied in the field of computer vision, in which the local features are first mapped to a codebook produced by clustering method and then represented by histogram of the words. One of drawbacks in BoW model is that the orderless histogram ignores the valuable spatial relationships among the features. In this study, we propose a novel framework based on a topographic independent component analysis (TICA), which enables the geometrically nearby feature components to be grouped together thereby bridge the semantic gap in BoW model. In addition, the compact feature obtained from TICA helps to build an efficient codebook. Furthermore, we introduce a new closeness measurement based on Neighbourhood Counting Measure (NCM) to improve the k Nearest Neighbour classification. The preliminary results based on KTH and Trecvid data demonstrate the proposed TICA/NCM approach increases the recognition accuracy and improve the efficiency of BoW model.
LanguageEnglish
Title of host publicationUnknown Host Publication
Number of pages9
Publication statusPublished - 2 Dec 2014
EventUCAmI 2014 -
Duration: 2 Dec 2014 → …

Conference

ConferenceUCAmI 2014
Period2/12/14 → …

Fingerprint

Independent component analysis
Computer vision
Semantics

Cite this

@inproceedings{c100784ebfeb481596fb1b5bdf8f32fa,
title = "Recognition by Enhanced Bag of Words Model via Topographic ICA",
abstract = "The Bag-of-Words (BoW) model has been increasingly applied in the field of computer vision, in which the local features are first mapped to a codebook produced by clustering method and then represented by histogram of the words. One of drawbacks in BoW model is that the orderless histogram ignores the valuable spatial relationships among the features. In this study, we propose a novel framework based on a topographic independent component analysis (TICA), which enables the geometrically nearby feature components to be grouped together thereby bridge the semantic gap in BoW model. In addition, the compact feature obtained from TICA helps to build an efficient codebook. Furthermore, we introduce a new closeness measurement based on Neighbourhood Counting Measure (NCM) to improve the k Nearest Neighbour classification. The preliminary results based on KTH and Trecvid data demonstrate the proposed TICA/NCM approach increases the recognition accuracy and improve the efficiency of BoW model.",
author = "Min Jing and Hui Wang and Kathy Clawson and SA Coleman and Shuwei Chen and Jun Liu and Bryan Scotney",
year = "2014",
month = "12",
day = "2",
language = "English",
booktitle = "Unknown Host Publication",

}

Jing, M, Wang, H, Clawson, K, Coleman, SA, Chen, S, Liu, J & Scotney, B 2014, Recognition by Enhanced Bag of Words Model via Topographic ICA. in Unknown Host Publication. UCAmI 2014, 2/12/14.

Recognition by Enhanced Bag of Words Model via Topographic ICA. / Jing, Min; Wang, Hui; Clawson, Kathy; Coleman, SA; Chen, Shuwei; Liu, Jun; Scotney, Bryan.

Unknown Host Publication. 2014.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Recognition by Enhanced Bag of Words Model via Topographic ICA

AU - Jing, Min

AU - Wang, Hui

AU - Clawson, Kathy

AU - Coleman, SA

AU - Chen, Shuwei

AU - Liu, Jun

AU - Scotney, Bryan

PY - 2014/12/2

Y1 - 2014/12/2

N2 - The Bag-of-Words (BoW) model has been increasingly applied in the field of computer vision, in which the local features are first mapped to a codebook produced by clustering method and then represented by histogram of the words. One of drawbacks in BoW model is that the orderless histogram ignores the valuable spatial relationships among the features. In this study, we propose a novel framework based on a topographic independent component analysis (TICA), which enables the geometrically nearby feature components to be grouped together thereby bridge the semantic gap in BoW model. In addition, the compact feature obtained from TICA helps to build an efficient codebook. Furthermore, we introduce a new closeness measurement based on Neighbourhood Counting Measure (NCM) to improve the k Nearest Neighbour classification. The preliminary results based on KTH and Trecvid data demonstrate the proposed TICA/NCM approach increases the recognition accuracy and improve the efficiency of BoW model.

AB - The Bag-of-Words (BoW) model has been increasingly applied in the field of computer vision, in which the local features are first mapped to a codebook produced by clustering method and then represented by histogram of the words. One of drawbacks in BoW model is that the orderless histogram ignores the valuable spatial relationships among the features. In this study, we propose a novel framework based on a topographic independent component analysis (TICA), which enables the geometrically nearby feature components to be grouped together thereby bridge the semantic gap in BoW model. In addition, the compact feature obtained from TICA helps to build an efficient codebook. Furthermore, we introduce a new closeness measurement based on Neighbourhood Counting Measure (NCM) to improve the k Nearest Neighbour classification. The preliminary results based on KTH and Trecvid data demonstrate the proposed TICA/NCM approach increases the recognition accuracy and improve the efficiency of BoW model.

M3 - Conference contribution

BT - Unknown Host Publication

ER -