京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第22800号 / 情博第730号 / 新制||情||125(附属図書館) / 京都大学大学院情報学研究科知能情報学専攻 / (主査)准教授 中澤 篤志, 教授 西野 恒, 教授 鹿島 久嗣 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DGAM

Urban bicycling is a big growing trend between men and women across the world. People seem to have become more aware of the numerous benefits of riding a bike, as well as more health oriented and environmentally friendly. Sweden is one of the countries with so called strong bicycle culture.  You can ride a bicycle almost anywhere, at any time of the year, and without spending a fortune. Cycling builds strength in a holistic manner since every single part of the body is involved in cycling. I choose to look into women cycling habit in Sweden. With my design I want to encourage woman’s cycling activity by offering comfortable design solutions to their clothing. The project focuses on the active years of a woman’s life, in average the age of 20- 45. The seasons considered are spring- autumn with the temperature range between 20°-5°C. The final design solutions will be applicable in the everyday life of a woman that is both looking to look fashionable and comfortable on the bike. The project intends to deliver a capsule collection where 3 to 5 outfits will be translated into 3D, prototyped and manufactured. One main reason for investigating such a matter is today’s very widely explored awareness of sustainability.  The project aims to encourage more sustainable design processes for the fashion market as well as eco lifestyle for women who choose to ride bikes instead of cars more often. It will propose a new possible design development process in which the user is more involved and can give some input, that way delivering a needed, better fitted and better sold product. The new design solution should satisfy the target group with a quality product that is not only comfortable and functional but also lasting for a long time. The final design will of course consider the esthetical feminine sophisticated look. Through this project I am looking to bring more comfort in the movement and functional detailing to non-sportswear. I want to investigate something apart from traditional sportswear, something that has not been considered much before; to incorporate functional sporty features and details to a more casual wear. I want to make use of special fabrics and develop new cuts, pockets, pleats and adjust patterns so that the final prototype would feel sporty but will look casual. Emphasis will be placed on developing understanding of fabric characteristics, accessing sources and knowledge and understanding of technological developments to inform innovative design. The purpose of the project is to develop a heightened awareness of the design methodologies available today that can be applied in fashion industry. This thesis will incorporate the traditional design development process applied in fashion and the user centered design method applied mostly in service and product design. Through research and prototyping, the aim will be to deliver a design that has been developed considering in a higher range user’s needs and preferences. User centred design is mainly looking to understand the user’s needs, wants and desires in order to improve the new design and human experience. When it comes to fashion, people are all individuals; they all have different needs and preferences. It is nearly impossible to try and satisfy them all, especially in mass produced fashions however the design development should in a way or another allow some user input. As little as suggestions on detailing, finishing features comfort and fit preferences could make a great difference for the success of the new design. Choosing to do a fashion project I initially wanted to investigate the problems with bad fit and uncomfortable women’s wear designs nowadays. With that in mind I was still missing a context, which clothes exactly and why. This is when I started to investigate specifically what women wear every day when riding bikes in the cities. Together with the city cycling trend, grows the demand for more comfort and function in the everyday clothes.  Clothing that can stay true to both performance and style is in big demand today as with clothes that look good both on and off the bike, there’s no need to have to change out of your cycling gear when you arrive at your destination. You can ride to work feeling comfortable and leave the office in style in the same set of clothes.

As e-commerce and online shopping have increased worldwide, the interest and research of intelligent fashion systems have expanded. Given the competitive nature of the fashion market business, digital marketplaces depend on determining customer preferences. The fashion preferences of the next generation of consumers, Generation Z, are highly discovered on social media, where new fashion styles have emerged. For digital marketplaces to gain the attraction of Generation Z consumers, an understanding of their fashion style preferences may be crucial. However, fashion style recognition remains challenging due to the subjective nature of fashion styles. Previous research has approached the task by fine-tuning pre-trained convolutional neural networks (CNNs). The disadvantage of this approach is that a CNN leveraged on its own fails to find subtle visual differences between clothing items. Hence, this thesis seeks to approach the clothing style recognition task as a fine-grained image recognition task by incorporating a component that allows the model to focus on specific parts of the input images, referred to as an attention mechanism, into the network. Specifically, a convolutional block attention module (CBAM) is added to a CNN. Based on the results, it is concluded that the fine-tuned CNN without the attention module achieves superior performance. In contrast, qualitative analysis conducted on GradCAM visualizations shows that the attention mechanism aids the CNN in capturing discriminative features, while the network without the attention module tends to make predictions based on dataset bias. For a fair comparison, future work should involve extending this research by refining the dataset or using an additional dataset. / I takt med att e-handel har ökat världen över har intresset och forskningen för intelligenta modesystem ökat. Modemarknadens konkurrenskraft har gjort digitala marknadsplatser beroende av att bestämma deras kunders preferenser. Modepreferenserna för nästa generations konsumenter, Generation Z, upptäcks ofta på sociala medier, där nya klädstilar har skapats. För att digitala marknadsplatser ska kunna locka Generation Z kan en förståelse för deras klädstilpreferenser vara avgörande. Igenkänning av klädstilar är dock fortfarande svårt på grund av klädtilars subjektiva natur. Tidigare forskning har finjusterat faltningsnätverk. Nackdelen med detta tillvägagångssätt är att ett faltningsnätverk som utnyttjas på egen hand inte lyckas hitta dem subtila visuella skillnader mellan klädesplagg. Därför definierar denna avhandling problemet som finkornig bildigenkänning genom att addera en komponent som gör att modellen kan fokusera på specifika delar av bilderna, kallad en uppmärksamhetsmekanism, i nätverket. Specifikt läggs en convolutional block attention module (CBAM) till i arkitekturen av ett faltningsnätverk. Baserat på resultaten dras slutsatsen att det finjusterade faltningsnätverket utan uppmärksamhetsmekanismen uppnår överlägsen prestanda. Däremot visar kvalitativ analys utförd på Grad-CAMvisualiseringar att uppmärksamhetsmekanismen hjälper faltningsnätverket att fokusera på de diskriminerande egenskaperna, medan nätverket utan uppmärksamhetsmekanismen tenderar att klassificera baserat på bias i inputdatan. För en rättvis jämförelse bör framtida arbete innebära ett förfinande av datamängden eller använda en ytterligare datamängd.