Silhouette Classification of Designer’s Collections in Luxury Fashion Brands

Abstract

The silhouette is an important element of fashion design. The designer's sensibility is expressed by the design changes in silhouette lines. The purpose of this research is to construct a silhouette classification criterion and renew the silhouette category, The analyzed images that capture the silhouette transformations were selected from recent designer's collections of luxury brands: Dior, a traditional luxury brand; Sacai, a new Japanese brand; and Doris Van Noten, one of the Antwerp Six. 11 measurement points used for the catwalk analysis were based on the reference points measured during the production of the clothes. The analysis utilized a combination of cluster analysis and multidimensional scaling. As the results, nine silhouette categories were obtained and the trend change was grasped visually by systematizing and classifying silhouettes that are important in predicting the future of fashion design. Moreover, this method could be applicable to automatic silhouette classification for unknown silhouettes.

Estimation of Garment Impression using Regression Models with Design Parameters and Image Features

“…A garment's appearance feature is considered essential for recommendation (He and McAuley, 2016;Tsuru et al, 2021). Much research has thus focused on recommendations using garment images (Yu et al, 2018;Guan et al, 2019).…”