Unsupervised generation of fashion editorials using deep generative model

IF 2.3 4区 管理学 Q1 MATERIALS SCIENCE, TEXTILES
Minjoo Kang, Jongsun Kim, Sungmin Kim
{"title":"Unsupervised generation of fashion editorials using deep generative model","authors":"Minjoo Kang,&nbsp;Jongsun Kim,&nbsp;Sungmin Kim","doi":"10.1186/s40691-023-00367-3","DOIUrl":null,"url":null,"abstract":"<div><p>This research intended to establish a new fashion-related artificial intelligence research topic concerning fashion editorials which could induce streams of further studies. A new fashion editorial dataset, which is a prerequisite in training an AI model, has been established in this study to meet the research purpose. A total of over 150K fashion editorials were initially collected and processed to satisfy necessary dataset conditions. A novel dataset of fashion editorials consisting of approximately 60K editorials is proposed through the process. In order to prove the adequacy of the new dataset, data distribution was analyzed and a generative model was selected and trained to attest that new fashion editorials can be created with the proposed editorial dataset. The results generated by the trained model were qualitatively investigated. The model has shown to have learned various features that compose editorials with the dataset, successfully generating fashion editorials. Quantitative evaluation with FID scores was conducted to support the selection of the generative model used for the qualitative assessment.</p></div>","PeriodicalId":555,"journal":{"name":"Fashion and Textiles","volume":"11 1","pages":""},"PeriodicalIF":2.3000,"publicationDate":"2024-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://fashionandtextiles.springeropen.com/counter/pdf/10.1186/s40691-023-00367-3","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Fashion and Textiles","FirstCategoryId":"88","ListUrlMain":"https://link.springer.com/article/10.1186/s40691-023-00367-3","RegionNum":4,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"MATERIALS SCIENCE, TEXTILES","Score":null,"Total":0}
引用次数: 0

Abstract

This research intended to establish a new fashion-related artificial intelligence research topic concerning fashion editorials which could induce streams of further studies. A new fashion editorial dataset, which is a prerequisite in training an AI model, has been established in this study to meet the research purpose. A total of over 150K fashion editorials were initially collected and processed to satisfy necessary dataset conditions. A novel dataset of fashion editorials consisting of approximately 60K editorials is proposed through the process. In order to prove the adequacy of the new dataset, data distribution was analyzed and a generative model was selected and trained to attest that new fashion editorials can be created with the proposed editorial dataset. The results generated by the trained model were qualitatively investigated. The model has shown to have learned various features that compose editorials with the dataset, successfully generating fashion editorials. Quantitative evaluation with FID scores was conducted to support the selection of the generative model used for the qualitative assessment.

利用深度生成模型无监督生成时尚社论
本研究旨在建立一个新的与时尚相关的人工智能研究课题,该课题涉及时尚社论,可引发进一步的研究。为了达到研究目的,本研究建立了一个新的时尚社论数据集,这是训练人工智能模型的先决条件。为了满足必要的数据集条件,我们初步收集并处理了超过 15 万篇时尚社论。在此过程中,我们提出了一个由大约 60K 篇社论组成的新型时尚社论数据集。为了证明新数据集的适当性,对数据分布进行了分析,并选择和训练了一个生成模型,以证明新的时尚社论可以通过建议的社论数据集创建。对训练模型生成的结果进行了定性研究。结果表明,该模型学会了利用数据集组成社论的各种特征,成功生成了时尚社论。利用 FID 分数进行了定量评估,以支持用于定性评估的生成模型的选择。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Fashion and Textiles
Fashion and Textiles Business, Management and Accounting-Marketing
CiteScore
4.40
自引率
4.20%
发文量
37
审稿时长
13 weeks
期刊介绍: Fashion and Textiles aims to advance knowledge and to seek new perspectives in the fashion and textiles industry worldwide. We welcome original research articles, reviews, case studies, book reviews and letters to the editor. The scope of the journal includes the following four technical research divisions: Textile Science and Technology: Textile Material Science and Technology; Dyeing and Finishing; Smart and Intelligent Textiles Clothing Science and Technology: Physiology of Clothing/Textile Products; Protective clothing ; Smart and Intelligent clothing; Sportswear; Mass customization ; Apparel manufacturing Economics of Clothing and Textiles/Fashion Business: Management of the Clothing and Textiles Industry; Merchandising; Retailing; Fashion Marketing; Consumer Behavior; Socio-psychology of Fashion Fashion Design and Cultural Study on Fashion: Aesthetic Aspects of Fashion Product or Design Process; Textiles/Clothing/Fashion Design; Fashion Trend; History of Fashion; Costume or Dress; Fashion Theory; Fashion journalism; Fashion exhibition.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信