The iMaterialist Fashion Attribute Dataset

Research output: Working paper › Preprint › Research

Standard

The iMaterialist Fashion Attribute Dataset. / Belongie, Serge; Guo, Sheng; Huang, Weilin; Zhang, Xiao; Srikhanta, Prasanna; Cui, Yin; Li, Yuan; Scott, Matthew R.; Adam, Hartwig.

arxiv.org, 2019.

Research output: Working paper › Preprint › Research

Harvard

Belongie, S, Guo, S, Huang, W, Zhang, X, Srikhanta, P, Cui, Y, Li, Y, Scott, MR & Adam, H 2019 'The iMaterialist Fashion Attribute Dataset' arxiv.org. <https://vision.cornell.edu/se3/wp-content/uploads/2019/06/1906.05750.pdf>

APA

Belongie, S., Guo, S., Huang, W., Zhang, X., Srikhanta, P., Cui, Y., Li, Y., Scott, M. R., & Adam, H. (2019). The iMaterialist Fashion Attribute Dataset. arxiv.org. https://vision.cornell.edu/se3/wp-content/uploads/2019/06/1906.05750.pdf

Vancouver

Belongie S, Guo S, Huang W, Zhang X, Srikhanta P, Cui Y et al. The iMaterialist Fashion Attribute Dataset. arxiv.org. 2019 Jun 13.

Author

Belongie, Serge ; Guo, Sheng ; Huang, Weilin ; Zhang, Xiao ; Srikhanta, Prasanna ; Cui, Yin ; Li, Yuan ; Scott, Matthew R. ; Adam, Hartwig. / The iMaterialist Fashion Attribute Dataset. arxiv.org, 2019.

Bibtex

@techreport{a14d8949d17545cda9c2ab632a14aa93,
title = "The iMaterialist Fashion Attribute Dataset",
abstract = "Large-scale image databases such as ImageNet have significantly advanced image classification and other visual recognition tasks. However much of these datasets are constructed only for single-label and coarse object-level classification. For real-world applications, multiple labels and fine-grained categories are often needed, yet very few such datasets exist publicly, especially those of large-scale and high quality. In this work, we contribute to the community a new dataset called iMaterialist Fashion Attribute (iFashion-Attribute) to address this problem in the fashion domain. The dataset is constructed from over one million fashion images with a label space that includes 8 groups of 228 fine-grained attributes in total. Each image is annotated by experts with multiple, high-quality fashion attributes. The result is the first known million-scale multi-label and fine-grained image dataset. We conduct extensive experiments and provide baseline results with modern deep Convolutional Neural Networks (CNNs). Additionally, we demonstrate models pre-trained on iFashion-Attribute achieve superior transfer learning performance on fashion related tasks compared with pre-training from ImageNet or other fashion datasets. Data is available at: this https URL",
author = "Serge Belongie and Sheng Guo and Weilin Huang and Xiao Zhang and Prasanna Srikhanta and Yin Cui and Yuan Li and Scott, {Matthew R.} and Hartwig Adam",
year = "2019",
month = jun,
day = "13",
language = "English",
publisher = "arxiv.org",
type = "WorkingPaper",
institution = "arxiv.org",

}

RIS

TY - UNPB

T1 - The iMaterialist Fashion Attribute Dataset

AU - Belongie, Serge

AU - Guo, Sheng

AU - Huang, Weilin

AU - Zhang, Xiao

AU - Srikhanta, Prasanna

AU - Cui, Yin

AU - Li, Yuan

AU - Scott, Matthew R.

AU - Adam, Hartwig

PY - 2019/6/13

Y1 - 2019/6/13

N2 - Large-scale image databases such as ImageNet have significantly advanced image classification and other visual recognition tasks. However much of these datasets are constructed only for single-label and coarse object-level classification. For real-world applications, multiple labels and fine-grained categories are often needed, yet very few such datasets exist publicly, especially those of large-scale and high quality. In this work, we contribute to the community a new dataset called iMaterialist Fashion Attribute (iFashion-Attribute) to address this problem in the fashion domain. The dataset is constructed from over one million fashion images with a label space that includes 8 groups of 228 fine-grained attributes in total. Each image is annotated by experts with multiple, high-quality fashion attributes. The result is the first known million-scale multi-label and fine-grained image dataset. We conduct extensive experiments and provide baseline results with modern deep Convolutional Neural Networks (CNNs). Additionally, we demonstrate models pre-trained on iFashion-Attribute achieve superior transfer learning performance on fashion related tasks compared with pre-training from ImageNet or other fashion datasets. Data is available at: this https URL

AB - Large-scale image databases such as ImageNet have significantly advanced image classification and other visual recognition tasks. However much of these datasets are constructed only for single-label and coarse object-level classification. For real-world applications, multiple labels and fine-grained categories are often needed, yet very few such datasets exist publicly, especially those of large-scale and high quality. In this work, we contribute to the community a new dataset called iMaterialist Fashion Attribute (iFashion-Attribute) to address this problem in the fashion domain. The dataset is constructed from over one million fashion images with a label space that includes 8 groups of 228 fine-grained attributes in total. Each image is annotated by experts with multiple, high-quality fashion attributes. The result is the first known million-scale multi-label and fine-grained image dataset. We conduct extensive experiments and provide baseline results with modern deep Convolutional Neural Networks (CNNs). Additionally, we demonstrate models pre-trained on iFashion-Attribute achieve superior transfer learning performance on fashion related tasks compared with pre-training from ImageNet or other fashion datasets. Data is available at: this https URL

UR - https://arxiv.org/abs/1906.05750

M3 - Preprint

BT - The iMaterialist Fashion Attribute Dataset

PB - arxiv.org

ER -

ID: 304511757