Sparse feature extraction model with Independent Subspace Analysis

Radhika Nath; Manju Manjunathaiah

doi:10.1007/978-3-030-13709-0_42

Sparse feature extraction model with Independent Subspace Analysis

Radhika Nath, Manju Manjunathaiah

Centre for Computer Science and Informatics Research

Research output: Chapter in Book/Report/Conference proceeding › Chapter (peer-reviewed) › peer-review

5 Downloads (Pure)

Abstract

Recent advances in deep learning models have demonstrated remarkable accuracy in object classification. However, the limitations of Convolutional Neural Networks such as the requirement for a large collection of labeled data for training and supervised learning process has called for enhanced feature representation and for unsupervised models.

In this paper we propose a novel unsupervised sparsity-based model using Independent Subspace Analysis (ISA) to implement a hierarchical network for feature extraction. The results of our empirical evaluation demonstrates an improved classification accuracy when max pooling is paired with square pooling within each layer. In addition to accuracy, we further show that it also reduces the data dimensions within the layers outperforming known sparsity-based models.

Original language	English
Title of host publication	Machine Learning, Optimization, and Data Science
Editors	G Nicosia , P Pardalos, G Giuffrida, R Umeton, V Sciacca
Publisher	Springer Nature Link
Chapter	42
Pages	494-505
Number of pages	12
ISBN (Electronic)	978-3-030-13709-0
ISBN (Print)	978-3-030-13708-3
DOIs	https://doi.org/10.1007/978-3-030-13709-0_42
Publication status	Published - 14 Feb 2019
Event	4th International Conference: LOD 2018 - Volterra, Italy Duration: 13 Sept 2018 → 16 Sept 2018

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	11331

Conference

Conference	4th International Conference
Country/Territory	Italy
City	Volterra
Period	13/09/18 → 16/09/18

Keywords

Biologically inspired vision models
Convolutional neural networks
Deep learning
Sparse models

Access to Document

10.1007/978-3-030-13709-0_42Licence: Other

Cite this

@inbook{93224c64ffaa49d7b0d8d5265965e9fc,

title = "Sparse feature extraction model with Independent Subspace Analysis",

abstract = "Recent advances in deep learning models have demonstrated remarkable accuracy in object classification. However, the limitations of Convolutional Neural Networks such as the requirement for a large collection of labeled data for training and supervised learning process has called for enhanced feature representation and for unsupervised models.In this paper we propose a novel unsupervised sparsity-based model using Independent Subspace Analysis (ISA) to implement a hierarchical network for feature extraction. The results of our empirical evaluation demonstrates an improved classification accuracy when max pooling is paired with square pooling within each layer. In addition to accuracy, we further show that it also reduces the data dimensions within the layers outperforming known sparsity-based models.",

keywords = "Biologically inspired vision models, Convolutional neural networks, Deep learning, Sparse models",

author = "Radhika Nath and Manju Manjunathaiah",

note = "{\textcopyright} Springer Nature Switzerland AG 2019; 4th International Conference : LOD 2018 ; Conference date: 13-09-2018 Through 16-09-2018",

year = "2019",

month = feb,

day = "14",

doi = "10.1007/978-3-030-13709-0_42",

language = "English",

isbn = "978-3-030-13708-3",

series = " Lecture Notes in Computer Science ",

publisher = "Springer Nature Link",

pages = "494--505",

editor = "{Nicosia }, G and P Pardalos and G Giuffrida and R Umeton and V Sciacca",

booktitle = "Machine Learning, Optimization, and Data Science",

address = "Netherlands",

}

Nath, R & Manjunathaiah, M 2019, Sparse feature extraction model with Independent Subspace Analysis. in G Nicosia , P Pardalos, G Giuffrida, R Umeton & V Sciacca (eds), Machine Learning, Optimization, and Data Science. Lecture Notes in Computer Science , vol. 11331, Springer Nature Link, pp. 494-505, 4th International Conference, Volterra, Italy, 13/09/18. https://doi.org/10.1007/978-3-030-13709-0_42

Sparse feature extraction model with Independent Subspace Analysis. / Nath, Radhika; Manjunathaiah, Manju.
Machine Learning, Optimization, and Data Science. ed. / G Nicosia ; P Pardalos; G Giuffrida; R Umeton; V Sciacca. Springer Nature Link, 2019. p. 494-505 ( Lecture Notes in Computer Science ; Vol. 11331).

Research output: Chapter in Book/Report/Conference proceeding › Chapter (peer-reviewed) › peer-review

TY - CHAP

T1 - Sparse feature extraction model with Independent Subspace Analysis

AU - Nath, Radhika

AU - Manjunathaiah, Manju

N1 - © Springer Nature Switzerland AG 2019

PY - 2019/2/14

Y1 - 2019/2/14

N2 - Recent advances in deep learning models have demonstrated remarkable accuracy in object classification. However, the limitations of Convolutional Neural Networks such as the requirement for a large collection of labeled data for training and supervised learning process has called for enhanced feature representation and for unsupervised models.In this paper we propose a novel unsupervised sparsity-based model using Independent Subspace Analysis (ISA) to implement a hierarchical network for feature extraction. The results of our empirical evaluation demonstrates an improved classification accuracy when max pooling is paired with square pooling within each layer. In addition to accuracy, we further show that it also reduces the data dimensions within the layers outperforming known sparsity-based models.

AB - Recent advances in deep learning models have demonstrated remarkable accuracy in object classification. However, the limitations of Convolutional Neural Networks such as the requirement for a large collection of labeled data for training and supervised learning process has called for enhanced feature representation and for unsupervised models.In this paper we propose a novel unsupervised sparsity-based model using Independent Subspace Analysis (ISA) to implement a hierarchical network for feature extraction. The results of our empirical evaluation demonstrates an improved classification accuracy when max pooling is paired with square pooling within each layer. In addition to accuracy, we further show that it also reduces the data dimensions within the layers outperforming known sparsity-based models.

KW - Biologically inspired vision models

KW - Convolutional neural networks

KW - Deep learning

KW - Sparse models

UR - http://www.scopus.com/inward/record.url?scp=85063591695&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-13709-0_42

DO - 10.1007/978-3-030-13709-0_42

M3 - Chapter (peer-reviewed)

SN - 978-3-030-13708-3

T3 - Lecture Notes in Computer Science

SP - 494

EP - 505

BT - Machine Learning, Optimization, and Data Science

A2 - Nicosia , G

A2 - Pardalos, P

A2 - Giuffrida, G

A2 - Umeton, R

A2 - Sciacca, V

PB - Springer Nature Link

T2 - 4th International Conference

Y2 - 13 September 2018 through 16 September 2018

ER -

Sparse feature extraction model with Independent Subspace Analysis

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this