Image classification of fine-grained fashion image based on style using pre-trained convolutional neural network

Yian Seo; Kyung Shik Shin

doi:10.1109/ICBDA.2018.8367713

Image classification of fine-grained fashion image based on style using pre-trained convolutional neural network

Yian Seo, Kyung Shik Shin

Division of Business Administration

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

25 Scopus citations

Abstract

Deep learning has emerged as a new methodology with continuous interests in artificial intelligence, and it can be applied in various business fields for better performance. In fashion business, deep learning, especially Convolutional Neural Network (CNN), is used in classification of apparel image. However, apparel classification can be difficult due to various apparel categories and lack of labeled image data for each category. Therefore, we propose to pre-train the GoogLeNet architecture on ImageNet dataset and fine-tune on our fine-grained fashion dataset based on design attributes. This will complement the small size of dataset and reduce the training time. After 10-fold experiments, the average final test accuracy results 62%.

Original language	English
Title of host publication	2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	387-390
Number of pages	4
ISBN (Electronic)	9781538647936
DOIs	https://doi.org/10.1109/ICBDA.2018.8367713
State	Published - 25 May 2018
Event	3rd IEEE International Conference on Big Data Analysis, ICBDA 2018 - Shanghai, China Duration: 9 Mar 2018 → 12 Mar 2018

Publication series

Name	2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018

Conference

Conference	3rd IEEE International Conference on Big Data Analysis, ICBDA 2018
Country/Territory	China
City	Shanghai
Period	9/03/18 → 12/03/18

Keywords

Convolutional Neural Network
fashion image
fine-grained classification
pre-trained network

Access to Document

10.1109/ICBDA.2018.8367713

Cite this

Seo, Y., & Shin, K. S. (2018). Image classification of fine-grained fashion image based on style using pre-trained convolutional neural network. In 2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018 (pp. 387-390). (2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICBDA.2018.8367713

@inproceedings{9206f6bcd01b404c86a9481f55162e9d,

title = "Image classification of fine-grained fashion image based on style using pre-trained convolutional neural network",

abstract = "Deep learning has emerged as a new methodology with continuous interests in artificial intelligence, and it can be applied in various business fields for better performance. In fashion business, deep learning, especially Convolutional Neural Network (CNN), is used in classification of apparel image. However, apparel classification can be difficult due to various apparel categories and lack of labeled image data for each category. Therefore, we propose to pre-train the GoogLeNet architecture on ImageNet dataset and fine-tune on our fine-grained fashion dataset based on design attributes. This will complement the small size of dataset and reduce the training time. After 10-fold experiments, the average final test accuracy results 62%.",

keywords = "Convolutional Neural Network, fashion image, fine-grained classification, pre-trained network",

author = "Yian Seo and Shin, {Kyung Shik}",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; null ; Conference date: 09-03-2018 Through 12-03-2018",

year = "2018",

month = may,

day = "25",

doi = "10.1109/ICBDA.2018.8367713",

language = "English",

series = "2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018",

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

pages = "387--390",

booktitle = "2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018",

}

Seo, Y & Shin, KS 2018, Image classification of fine-grained fashion image based on style using pre-trained convolutional neural network. in 2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018. 2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018, Institute of Electrical and Electronics Engineers Inc., pp. 387-390, 3rd IEEE International Conference on Big Data Analysis, ICBDA 2018, Shanghai, China, 9/03/18. https://doi.org/10.1109/ICBDA.2018.8367713

Image classification of fine-grained fashion image based on style using pre-trained convolutional neural network. / Seo, Yian; Shin, Kyung Shik.

2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 387-390 (2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018).

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

TY - GEN

T1 - Image classification of fine-grained fashion image based on style using pre-trained convolutional neural network

AU - Seo, Yian

AU - Shin, Kyung Shik

PY - 2018/5/25

Y1 - 2018/5/25

N2 - Deep learning has emerged as a new methodology with continuous interests in artificial intelligence, and it can be applied in various business fields for better performance. In fashion business, deep learning, especially Convolutional Neural Network (CNN), is used in classification of apparel image. However, apparel classification can be difficult due to various apparel categories and lack of labeled image data for each category. Therefore, we propose to pre-train the GoogLeNet architecture on ImageNet dataset and fine-tune on our fine-grained fashion dataset based on design attributes. This will complement the small size of dataset and reduce the training time. After 10-fold experiments, the average final test accuracy results 62%.

AB - Deep learning has emerged as a new methodology with continuous interests in artificial intelligence, and it can be applied in various business fields for better performance. In fashion business, deep learning, especially Convolutional Neural Network (CNN), is used in classification of apparel image. However, apparel classification can be difficult due to various apparel categories and lack of labeled image data for each category. Therefore, we propose to pre-train the GoogLeNet architecture on ImageNet dataset and fine-tune on our fine-grained fashion dataset based on design attributes. This will complement the small size of dataset and reduce the training time. After 10-fold experiments, the average final test accuracy results 62%.

KW - Convolutional Neural Network

KW - fashion image

KW - fine-grained classification

KW - pre-trained network

UR - http://www.scopus.com/inward/record.url?scp=85048490367&partnerID=8YFLogxK

U2 - 10.1109/ICBDA.2018.8367713

DO - 10.1109/ICBDA.2018.8367713

M3 - Conference contribution

AN - SCOPUS:85048490367

T3 - 2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018

SP - 387

EP - 390

BT - 2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 9 March 2018 through 12 March 2018

ER -

Seo Y, Shin KS. Image classification of fine-grained fashion image based on style using pre-trained convolutional neural network. In 2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 387-390. (2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018). https://doi.org/10.1109/ICBDA.2018.8367713

Image classification of fine-grained fashion image based on style using pre-trained convolutional neural network

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this