https://scholars.lib.ntu.edu.tw/handle/123456789/608030
標題: | From street photos to fashion trends: Leveraging user-provided noisy labels for fashion understanding | 作者: | Huang F.-H Lu H.-M YAO-WEN HSU HSIN-MIN LU |
關鍵字: | Deep learning;Fashion dataset;Fashion trends;Image clustering;Image recognition;Multi-label classification;Multi-task learning;Noisy labels;Forecasting;Social networking (online);Fashion styles;Feature vectors;Image clusters;Image datasets;Prediction performance;Share knowledge;Social Network Sites;Technical challenges;Image analysis | 公開日期: | 2021 | 卷: | 9 | 起(迄)頁: | 49189-49205 | 來源出版物: | IEEE Access | 摘要: | There is increased interest in using street photos to understand fashion trends. Though street photos usually contain rich clothing information, there are several technical challenges to their analysis. First, street photos collected from social media sites often contain user-provided noisy labels, and training models using these labels may deteriorate prediction performance. Second, most existing methods predict multiple clothing attributes individually and do not consider the potential to share knowledge between related tasks. In addition to these technical challenges, most fashion image datasets created by previous studies focus on American and European fashion styles. To address these technical challenges and understand fashion trends in Asia, we created RichWear, a new street fashion dataset containing 322,198 images with various text labels for fashion analysis. This dataset, collected from an Asian social network site, focuses on street styles in Japan and other Asian areas. RichWear provides a subset of expert-verified labels in addition to user-provided noisy labels for model training and evaluation. We propose the Fashion Attributes Recognition Network (FARNet) based on the multi-task learning framework to improve fashion recognition. Instead of predicting each clothing attribute individually, FARNet predicts three types of attributes simultaneously, and, once trained, this network leverages the noisy labels and generates corrected labels based on the input images. Experimental results show that this approach significantly outperforms existing methods. Applying the trained model to the RichWear dataset, we report Asian fashion trends and street styles based on predicted labels and image clusters from latent feature vectors. ? 2021 American Institute of Physics Inc.. All rights reserved. |
URI: | https://www.scopus.com/inward/record.uri?eid=2-s2.0-85103949125&doi=10.1109%2fACCESS.2021.3069245&partnerID=40&md5=c39657cc3d43473211a4abf61c1551f1 https://scholars.lib.ntu.edu.tw/handle/123456789/608030 |
ISSN: | 21693536 | DOI: | 10.1109/ACCESS.2021.3069245 |
顯示於: | 國際企業學系 |
在 IR 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。