Informative census transform for very low-resolution image representation

Sungmoon Jeong; Hosun Lee; Nak Young Chong

doi:10.1109/ROMAN.2014.6926291

Informative census transform for very low-resolution image representation

Sungmoon Jeong, Hosun Lee, Nak Young Chong

Japan Advanced Institute of Science and Technology

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

Abstract

Our paper newly presents unsupervised feature representation method for very low-resolution (VLR) images called informative census transform (ICT) based on statistical analysis of CT binary features and submodular optimization. A new cost function is designed to measure the informativeness of each binary feature: (1) an individual informativeness of features to represent unlabeled image dataset and (2) relative informativeness between binary features to represent different binary features. Therefore, we considered informativeness of binary feature according to two relationship (1) between feature space and image space, and (2) between different features within same feature space. Moreover, two constraints are designed by considering sub-modular characteristics to guarantee theoretical performance and fast optimization via simple greedy algorithm. Experimental results show that the proposed ICT features with two constraints outperforms the traditional CT features in terms of recognition performance and computational cost at VLR problem.

Original language	English
Title of host publication	IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication
Subtitle of host publication	Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions
Editors	Rui Loureiro, Aris Alissandrakis, Adriana Tapus, Selma Sabanovic, Fumihide Tanaka, Yukie Nagai
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	435-440
Number of pages	6
ISBN (Electronic)	9781479967636
DOIs	https://doi.org/10.1109/ROMAN.2014.6926291
State	Published - 15 Oct 2014
Event	23rd IEEE International Symposium on Robot and Human Interactive Communication, IEEE RO-MAN 2014 - Edinburgh, United Kingdom Duration: 25 Aug 2014 → 29 Aug 2014

Publication series

Name	IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions

Conference

Conference	23rd IEEE International Symposium on Robot and Human Interactive Communication, IEEE RO-MAN 2014
Country/Territory	United Kingdom
City	Edinburgh
Period	25/08/14 → 29/08/14

Access to Document

10.1109/ROMAN.2014.6926291

Cite this

Jeong, S., Lee, H., & Chong, N. Y. (2014). Informative census transform for very low-resolution image representation. In R. Loureiro, A. Alissandrakis, A. Tapus, S. Sabanovic, F. Tanaka, & Y. Nagai (Eds.), IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions (pp. 435-440). Article 6926291 (IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ROMAN.2014.6926291

Jeong, Sungmoon ; Lee, Hosun ; Chong, Nak Young. / Informative census transform for very low-resolution image representation. IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions. editor / Rui Loureiro ; Aris Alissandrakis ; Adriana Tapus ; Selma Sabanovic ; Fumihide Tanaka ; Yukie Nagai. Institute of Electrical and Electronics Engineers Inc., 2014. pp. 435-440 (IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions).

@inproceedings{2653d621ab4340208f5523cf2415482b,

title = "Informative census transform for very low-resolution image representation",

abstract = "Our paper newly presents unsupervised feature representation method for very low-resolution (VLR) images called informative census transform (ICT) based on statistical analysis of CT binary features and submodular optimization. A new cost function is designed to measure the informativeness of each binary feature: (1) an individual informativeness of features to represent unlabeled image dataset and (2) relative informativeness between binary features to represent different binary features. Therefore, we considered informativeness of binary feature according to two relationship (1) between feature space and image space, and (2) between different features within same feature space. Moreover, two constraints are designed by considering sub-modular characteristics to guarantee theoretical performance and fast optimization via simple greedy algorithm. Experimental results show that the proposed ICT features with two constraints outperforms the traditional CT features in terms of recognition performance and computational cost at VLR problem.",

author = "Sungmoon Jeong and Hosun Lee and Chong, {Nak Young}",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.; 23rd IEEE International Symposium on Robot and Human Interactive Communication, IEEE RO-MAN 2014 ; Conference date: 25-08-2014 Through 29-08-2014",

year = "2014",

month = oct,

day = "15",

doi = "10.1109/ROMAN.2014.6926291",

language = "English",

series = "IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "435--440",

editor = "Rui Loureiro and Aris Alissandrakis and Adriana Tapus and Selma Sabanovic and Fumihide Tanaka and Yukie Nagai",

booktitle = "IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication",

address = "United States",

}

Jeong, S, Lee, H & Chong, NY 2014, Informative census transform for very low-resolution image representation. in R Loureiro, A Alissandrakis, A Tapus, S Sabanovic, F Tanaka & Y Nagai (eds), IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions., 6926291, IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions, Institute of Electrical and Electronics Engineers Inc., pp. 435-440, 23rd IEEE International Symposium on Robot and Human Interactive Communication, IEEE RO-MAN 2014, Edinburgh, United Kingdom, 25/08/14. https://doi.org/10.1109/ROMAN.2014.6926291

Informative census transform for very low-resolution image representation. / Jeong, Sungmoon; Lee, Hosun; Chong, Nak Young.
IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions. ed. / Rui Loureiro; Aris Alissandrakis; Adriana Tapus; Selma Sabanovic; Fumihide Tanaka; Yukie Nagai. Institute of Electrical and Electronics Engineers Inc., 2014. p. 435-440 6926291 (IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions).

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

TY - GEN

T1 - Informative census transform for very low-resolution image representation

AU - Jeong, Sungmoon

AU - Lee, Hosun

AU - Chong, Nak Young

PY - 2014/10/15

Y1 - 2014/10/15

N2 - Our paper newly presents unsupervised feature representation method for very low-resolution (VLR) images called informative census transform (ICT) based on statistical analysis of CT binary features and submodular optimization. A new cost function is designed to measure the informativeness of each binary feature: (1) an individual informativeness of features to represent unlabeled image dataset and (2) relative informativeness between binary features to represent different binary features. Therefore, we considered informativeness of binary feature according to two relationship (1) between feature space and image space, and (2) between different features within same feature space. Moreover, two constraints are designed by considering sub-modular characteristics to guarantee theoretical performance and fast optimization via simple greedy algorithm. Experimental results show that the proposed ICT features with two constraints outperforms the traditional CT features in terms of recognition performance and computational cost at VLR problem.

AB - Our paper newly presents unsupervised feature representation method for very low-resolution (VLR) images called informative census transform (ICT) based on statistical analysis of CT binary features and submodular optimization. A new cost function is designed to measure the informativeness of each binary feature: (1) an individual informativeness of features to represent unlabeled image dataset and (2) relative informativeness between binary features to represent different binary features. Therefore, we considered informativeness of binary feature according to two relationship (1) between feature space and image space, and (2) between different features within same feature space. Moreover, two constraints are designed by considering sub-modular characteristics to guarantee theoretical performance and fast optimization via simple greedy algorithm. Experimental results show that the proposed ICT features with two constraints outperforms the traditional CT features in terms of recognition performance and computational cost at VLR problem.

UR - http://www.scopus.com/inward/record.url?scp=84937538604&partnerID=8YFLogxK

U2 - 10.1109/ROMAN.2014.6926291

DO - 10.1109/ROMAN.2014.6926291

M3 - Conference contribution

AN - SCOPUS:84937538604

T3 - IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions

SP - 435

EP - 440

BT - IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication

A2 - Loureiro, Rui

A2 - Alissandrakis, Aris

A2 - Tapus, Adriana

A2 - Sabanovic, Selma

A2 - Tanaka, Fumihide

A2 - Nagai, Yukie

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 23rd IEEE International Symposium on Robot and Human Interactive Communication, IEEE RO-MAN 2014

Y2 - 25 August 2014 through 29 August 2014

ER -

Jeong S, Lee H, Chong NY. Informative census transform for very low-resolution image representation. In Loureiro R, Alissandrakis A, Tapus A, Sabanovic S, Tanaka F, Nagai Y, editors, IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions. Institute of Electrical and Electronics Engineers Inc. 2014. p. 435-440. 6926291. (IEEE RO-MAN 2014 - 23rd IEEE International Symposium on Robot and Human Interactive Communication: Human-Robot Co-Existence: Adaptive Interfaces and Systems for Daily Life, Therapy, Assistance and Socially Engaging Interactions). doi: 10.1109/ROMAN.2014.6926291

Informative census transform for very low-resolution image representation

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this