How do Convolutional Neural Networks Learn Design?

Shailza Jolly, Brian Kenji Iwana, Ryohei Kuroki, Seiichi Uchida

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

Abstract

In this paper, we aim to understand the design principles in book cover images which are carefully crafted by experts. Book covers are designed in a unique way, specific to genres which convey important information to their readers. By using Convolutional Neural Networks (CNN) to predict book genres from cover images, visual cues which distinguish genres can be highlighted and analyzed. In order to understand these visual clues contributing towards the decision of a genre, we present the application of Layer-wise Relevance Propagation (LRP) on the book cover image classification results. We use LRP to explain the pixel-wise contributions of book cover design and highlight the design elements contributing towards particular genres. In addition, with the use of state-of-the-art object and text detection methods, insights about genre-specific book cover designs are discovered.

Original languageEnglish
Title of host publication2018 24th International Conference on Pattern Recognition, ICPR 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1085-1090
Number of pages6
ISBN (Electronic)9781538637883
DOIs
Publication statusPublished - Nov 26 2018
Event24th International Conference on Pattern Recognition, ICPR 2018 - Beijing, China
Duration: Aug 20 2018Aug 24 2018

Publication series

NameProceedings - International Conference on Pattern Recognition
Volume2018-August
ISSN (Print)1051-4651

Other

Other24th International Conference on Pattern Recognition, ICPR 2018
CountryChina
CityBeijing
Period8/20/188/24/18

Fingerprint

Neural networks
Image classification
Pixels

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

Cite this

Jolly, S., Iwana, B. K., Kuroki, R., & Uchida, S. (2018). How do Convolutional Neural Networks Learn Design? In 2018 24th International Conference on Pattern Recognition, ICPR 2018 (pp. 1085-1090). [8545624] (Proceedings - International Conference on Pattern Recognition; Vol. 2018-August). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICPR.2018.8545624

How do Convolutional Neural Networks Learn Design? / Jolly, Shailza; Iwana, Brian Kenji; Kuroki, Ryohei; Uchida, Seiichi.

2018 24th International Conference on Pattern Recognition, ICPR 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 1085-1090 8545624 (Proceedings - International Conference on Pattern Recognition; Vol. 2018-August).

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

Jolly, S, Iwana, BK, Kuroki, R & Uchida, S 2018, How do Convolutional Neural Networks Learn Design? in 2018 24th International Conference on Pattern Recognition, ICPR 2018., 8545624, Proceedings - International Conference on Pattern Recognition, vol. 2018-August, Institute of Electrical and Electronics Engineers Inc., pp. 1085-1090, 24th International Conference on Pattern Recognition, ICPR 2018, Beijing, China, 8/20/18. https://doi.org/10.1109/ICPR.2018.8545624
Jolly S, Iwana BK, Kuroki R, Uchida S. How do Convolutional Neural Networks Learn Design? In 2018 24th International Conference on Pattern Recognition, ICPR 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 1085-1090. 8545624. (Proceedings - International Conference on Pattern Recognition). https://doi.org/10.1109/ICPR.2018.8545624
Jolly, Shailza ; Iwana, Brian Kenji ; Kuroki, Ryohei ; Uchida, Seiichi. / How do Convolutional Neural Networks Learn Design?. 2018 24th International Conference on Pattern Recognition, ICPR 2018. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 1085-1090 (Proceedings - International Conference on Pattern Recognition).
@inproceedings{f05ccc25ba374cba815139dde486f30d,
title = "How do Convolutional Neural Networks Learn Design?",
abstract = "In this paper, we aim to understand the design principles in book cover images which are carefully crafted by experts. Book covers are designed in a unique way, specific to genres which convey important information to their readers. By using Convolutional Neural Networks (CNN) to predict book genres from cover images, visual cues which distinguish genres can be highlighted and analyzed. In order to understand these visual clues contributing towards the decision of a genre, we present the application of Layer-wise Relevance Propagation (LRP) on the book cover image classification results. We use LRP to explain the pixel-wise contributions of book cover design and highlight the design elements contributing towards particular genres. In addition, with the use of state-of-the-art object and text detection methods, insights about genre-specific book cover designs are discovered.",
author = "Shailza Jolly and Iwana, {Brian Kenji} and Ryohei Kuroki and Seiichi Uchida",
year = "2018",
month = "11",
day = "26",
doi = "10.1109/ICPR.2018.8545624",
language = "English",
series = "Proceedings - International Conference on Pattern Recognition",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "1085--1090",
booktitle = "2018 24th International Conference on Pattern Recognition, ICPR 2018",
address = "United States",

}

TY - GEN

T1 - How do Convolutional Neural Networks Learn Design?

AU - Jolly, Shailza

AU - Iwana, Brian Kenji

AU - Kuroki, Ryohei

AU - Uchida, Seiichi

PY - 2018/11/26

Y1 - 2018/11/26

N2 - In this paper, we aim to understand the design principles in book cover images which are carefully crafted by experts. Book covers are designed in a unique way, specific to genres which convey important information to their readers. By using Convolutional Neural Networks (CNN) to predict book genres from cover images, visual cues which distinguish genres can be highlighted and analyzed. In order to understand these visual clues contributing towards the decision of a genre, we present the application of Layer-wise Relevance Propagation (LRP) on the book cover image classification results. We use LRP to explain the pixel-wise contributions of book cover design and highlight the design elements contributing towards particular genres. In addition, with the use of state-of-the-art object and text detection methods, insights about genre-specific book cover designs are discovered.

AB - In this paper, we aim to understand the design principles in book cover images which are carefully crafted by experts. Book covers are designed in a unique way, specific to genres which convey important information to their readers. By using Convolutional Neural Networks (CNN) to predict book genres from cover images, visual cues which distinguish genres can be highlighted and analyzed. In order to understand these visual clues contributing towards the decision of a genre, we present the application of Layer-wise Relevance Propagation (LRP) on the book cover image classification results. We use LRP to explain the pixel-wise contributions of book cover design and highlight the design elements contributing towards particular genres. In addition, with the use of state-of-the-art object and text detection methods, insights about genre-specific book cover designs are discovered.

UR - http://www.scopus.com/inward/record.url?scp=85059741230&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85059741230&partnerID=8YFLogxK

U2 - 10.1109/ICPR.2018.8545624

DO - 10.1109/ICPR.2018.8545624

M3 - Conference contribution

AN - SCOPUS:85059741230

T3 - Proceedings - International Conference on Pattern Recognition

SP - 1085

EP - 1090

BT - 2018 24th International Conference on Pattern Recognition, ICPR 2018

PB - Institute of Electrical and Electronics Engineers Inc.

ER -