Systems approach to clothing for disabled users. Why is it difficult for disabled users to find suitable clothing

Today's clothing market is not adapted for people with unusual body dimensions and/or different kinds of functional impairments. They will be refered to as disabled users. The method of user-oriented product development combined with the soft systems methodology has been applied in order to understand and possibly solve their problem. Sixty five disabled users were interviewed. The results show that the disabled users must be divided into subgroups. Their requirements are very much dependent on their impairment not only regarding the fit and function of the clothing, but also regarding shopping and service in the stores. The symbolic values of clothing are just as important as the functional values. However, it is essential not to focus on the end product only, but to look at the problem as a system consisting of: manufacturing, marketing and end use. A new method of taking three dimensional measurements from range images of the human body has been developed. In combination with other developments in information technology it will be possible to synthesise a new system where the manufacturers and retailers of clothing can co-operate in order to give service to disabled users.     

Environmental mappings as a basis for the formulation of clothing demands

The basic functional demands on working clothes are protection and comfort. These basic demands are determined by the three variables: (1) level of activity of the worker, (2) environmental pollution and impacts, (3) surrounding climate. The environmental mapping is a systematic method to detect, quantify and analyse the demands on working clothes in different working situations. A pilot study was carried out in Swedish industry to investigate the general conditions regarding working clothes and to find the correspondence between clothing demands and environmental factors. The second study was motivated by clothing problems in the mechanical engineering industry. In both studies the three determining variables for the clothing of each interviewed person were given certain code values according to their measured or graded values. The pilot study confirmed the theoretical model for work classification. The model serves as an instrument for finding relevant clothing demands for clothing development and clothing procurement, and as a source of information for the workers so that they can check the clothing demands for their own working conditions. In the study of the mechanical engineering industry all 310 interviewed workers with 20 different professions could be divided into three groups with different clothing demands. These quantified functional clothing demands were transformed into certain demands in terms of textile material and model design. The development work finally resulted in new clothes, an overall and bib-and-brace trousers in various suitable materials.     

Clothing fasteners for long-term-care patients: evaluation of standard closures and prototypes on test garments

For the elderly long-term-care patient it is of great psychological and practical importance to be able to dress and undress without assistance. However, the design and position of clothing fasteners often lead to a restriction of the functional capacity of the patient. The aim of this study was to facilitate a more independent daily life for long-term care patients. New fasteners were developed in accordance with a user-oriented product development process where a detailed knowledge of the final users and their capacity and limitations, problems, needs and wishes form the basis for further development. In an initial experimental study (I) the practical use of different standard clothing fasteners (button, hook and eye, nylon pressure tape and zipper) was evaluated. Different fasteners and positions are favourable to different patient groups. Wrist angle in different hand positions is crucial for the function of fasteners for persons with reduced hand function. User demands on clothing fasteners were specified and prototypes developed. In a subsequent study (II) an adapted oval button and a 'finger strap' alternative to hook-and-eye were designed and evaluated, together with a standard button. The oval button in combination with a vertical buttonhole improved the function for most patients, and the front position for the fasteners was superior to a diagonal or lateral position on the chest. The finger strap was of advantage to patients with hemiplegia and joint complaints but was difficult for many of the subjects to understand, being a technical innovation.     

基于衣物共现信息与多任务学习的衣物识别*

多任务学习(MTL)未考虑先验概率对学习的影响.针对这一问题,文中提出基于衣物共现信息与多任务学习的衣物识别方法(CA-MLT),通过在MTL模型中加入先验约束项整合衣物共现信息,并对传统的扩展梯度算法进行相应的修改,从而提高衣物类别分类器的性能.实验表明,CA-MLT的平均性能优于单一任务学习、神经网络及传统的多任务学习等方法,训练结果便于可视化,可用于特征选择.     

Smart Clothing Prototype for the Arctic Environment

Continuous miniaturisation of electronic components has made it possible to create smaller and smaller electrical devices which can be worn and carried all the time. Together with developing fibre and textile technologies, this has enabled the creation of truly usable smart clothes that resemble clothes more than wearable computing equipment. These intelligent clothes are worn like ordinary clothing and provide help in various situations according to the application area. This paper describes the design and implementation of a survival smart clothing prototype for the arctic environment. Concept development, electrical design, and non-electrical features are discussed. The suit provides communication, positioning, and navigation aids for the user. Depending on the measurements of the human and the environment, the suit decides whether an emergency message should be sent. The user can control the system with a user interface called a Yo-Yo. The functionality of the suit has been tested in an arctic environment.     

Flexible shape control for automatic resizing of apparel products

We provide a flexible shape control technique in this paper for the automatic resizing of apparel products. The automatic resizing function has become an essential part of the 3D garment CAD systems to generate user customized apparel products for individuals with variant body shapes. The human bodies are usually represented by piecewise linear mesh surfaces with consistent connectivity. The shape of apparel products can then be warped from the space around a human body to the space around another body by computing the new positions of points on apparel products. However, one major limitation of this kind of automatic resizing technique is that the apparel products are always distorted along the shape of the human bodies. This is a required deformation for tight clothes but not an expected result for other types of clothes. To solve this problem, we investigate a method to preserve the shape of user-defined features on the apparel products. As the apparel products are often represented by discrete surfaces with non-manifold entities, the existing mesh processing approaches that preserve the local shape cannot be applied here. A new algorithm consisting of three steps is developed in this paper. First, the apparel product is warped from the reference human body to the space around the target human body. Second, the shape of features is optimized to match their original shape before the warping. Lastly, discrete surfaces of the apparel product are deformed again under an optimization framework to match their original shapes locally while interpolating the shape of features determined in the previous step.     

基于HOG 和几何特征的服装细节要素识别与分类

随着人们对服装产品多元化需求的增强,无论是对于经营者还是消费者,服装分 类都显得十分必要。现有方法大都基于服装整体做出处理,而忽略了服装细节要素的特征。提 出针对服装细节,如衣领类型、袖子和下装长度等进行识别分类方法。在轮廓提取的基础上, 针对衣领位置不确定、领口形状受周边花纹干扰等难点,设计了在多尺度HOG 结果上进行投 票的方法,并结合基于角点检测的几何特征提取如关键尺寸比例计算等,用SVM 完成训练分 类。最后利用多个特征搭配系数矩阵给出服装搭配建议。实验表明,该方法能够有效地完成上 述服装细节要素分类,对自动搭配推荐也有一定实用价值。     

How to Wear Beautifully? Clothing Pair Recommendation

In this paper, we present a practical system to automatically suggest the most pairing clothing items, given the reference clothing (upper-body or low-body). This has been a challenge, due to clothes having a variety of categories. Clothing is one of the most informative cues for human appearance. In our daily life, people need to wear properly and beautifully to show their confidence, politeness and social status in various occasions. However, it is not easy to decide what to wear and how to coordinate their own clothes. To address this problem, we propose a recommendation approach that includes clothing region detection, clothing pair recommendation and distance fusion. Clothing region detection based on Faster R-CNN is used to detect clothing region. Clothing pair recommendation consists of a quadruple network architecture, where one dual network of the architecture adopts Siamese convolution neural network architecture. Training examples are pairs of upper-body and low-body clothing items that are either compatible or incompatible. The other dual network is used to learn clothing style features of the input image. This framework is designed to learn a feature transformation from the images of clothing items into two latent spaces, which we call them compatible space and style space respectively. After training the two dual networks, we use a distance fusion method to fuse the features extracted from the compatible and style dual networks. To acquire an optimized model and verify our proposed method, we expand an existing large clothing dataset WoG (Weather-to-Garment), and the resulted dataset is called “How to Wear Beautifully” (H2WB). Experiments on the H2WB dataset demonstrate that our approach is effective with clothing region detection and clothing pair recommendation as well as distance fusion.     

Clothing-invariant gait identification using part-based clothing categorization and adaptive weight control

Variations in clothing alter an individual's appearance, making the problem of gait identification much more difficult. If the type of clothing differs between the gallery and a probe, certain parts of the silhouettes are likely to change and the ability to discriminate subjects decreases with respect to these parts. A part-based approach, therefore, has the potential of selecting the appropriate parts. This paper proposes a method for part-based gait identification in the light of substantial clothing variations. We divide the human body into eight sections, including four overlapping ones, since the larger parts have a higher discrimination capability, while the smaller parts are more likely to be unaffected by clothing variations. Furthermore, as there are certain clothes that are common to different parts, we present a categorization for items of clothing that groups similar clothes. Next, we exploit the discrimination capability as a matching weight for each part and control the weights adaptively based on the distribution of distances between the probe and all the galleries. The results of the experiments using our large-scale gait dataset with clothing variations show that the proposed method achieves far better performance than other approaches.     

Editorial: comfort and discomfort studies demonstrate the need for a new model

The term comfort is often seen relating to the marketing of products like chairs, cars, clothing, hand tools and even airplane tickets, while in the scientific literature, the term discomfort shows up often, since it is used in research. Few papers explain the concept of a localized comfort experience in relation to product use, although people use these products daily. Therefore, in this special issue, the concept of product comfort is studied further. In this editorial an overview of comfort models has been made, evaluated with the papers from the special issue and a new comfort/discomfort model is proposed to increase our understanding of the factors influencing comfort and discomfort experiences.     

Aesthetic and functional analysis for product model validation in reverse engineering applications

Reverse engineering techniques are broadly used in the development process of products with meaningful aesthetic properties. Many stylists prefer to evaluate the product shape on the basis of a full-scale hand made physical mock-up. Such model has then to be converted into a 3D CAD model, to begin the product engineering and production processes.A critical aspect of the RE process is that the physical mock-up, made by the stylist, usually does not take into account the engineering production constraints. The surface reconstruction activity must then be followed by a modelling phase, where the engineering designer modifies the reconstructed model shape in order to make it suitable for production.At the end of this process, no matter how accurate and precise the surface reconstruction phase has been, the product model will be different from the original mock-up, and then it needs to be submitted to the stylist for the validation of its shape.For the stylist, getting used to working with physical models, the shape validation on a virtual model is not a trivial task.The objective of this research is to develop a software tool to support the identification of aesthetic and functional regions of the product model shape, that have been modified with respect to the original mock-up.The proposed approach is based on the decomposition of a 3D surface analysis problem into a simpler 2D curves analysis problem. This approach simulates the traditional method used by stylists to evaluate the quality of shapes.The implementation of the developed algorithms has been performed using a commercial software package (I-deas FreeForm by EDS/Unigraphics) and it has been successfully applied on real test cases.     

论哥特式时期的建筑与服装的美学共通

建筑与服装同属于立体艺术,一个有凝固之美,一个有流动之美,而同一时期的建筑和服装的审美标准与指导准则总是和谐统一的。哥特式建筑与哥特式服装可以称得上是代表之作。哥特式艺术出现于12世纪的欧洲,那时中世纪艺术形式的发展虽然已经经历了深刻的变化,但是宗教神学的统治仍然高高在上。在这种时代背景下,哥特式建筑和哥特式服装应运而生,虽然表现形式与功能不同,但外部特征却有许多相似之处。     

User-oriented evaluation of mechanical single-channel axial pipettes

Hand tools should be designed so that they are comfortable to use, fit the hand and are user-oriented. Six different manual, single-channel axial pipettes were evaluated for such objective outcomes as muscular activity, wrist postures and efficiency, as well as for subjective outcomes concerning self-assessed features of pipette usability and musculoskeletal strain. Ten experienced laboratory employees volunteered for the study. The results showed that light and short pipettes with better tool comfort resulted in reduced muscular activity and perceived musculoskeletal strain when they were compared with a long and heavy pipette. There were no differences in the efficiency between the different pipettes. Combining both the objective and subjective measures enabled a broader evaluation of product usability. The results of this study can be used both in product development and as information on which to base the purchase of new pipettes for laboratory work.     

Inchworm robot grippers for clothes manipulation

Manipulating deformable objects like clothes, plastic, and paper by a robot is very challenging. This paper focuses on clothes manipulation as an example. A tracing manipulation method is used here to find a corner of the clothes. In this paper, tracing refers to tracing the clothes’ edge, with the robot’s movement based on feedback from sensors. One difficulty during this edge tracing is to make the robot trace smoothly and speedily without dropping the clothes in the process. This is due to the fact that deformable objects are sensitive to contact forces. A solution to this problem is to design a special robot gripper that can trace the clothes without having to worry about the clothes slipping away. In this paper, the development of inchworm-type grippers is proposed. Two sets of grippers inside a robot hand will allow the robot to trace the clothes more freely because there will always be a gripper firmly holding the clothes at any time. A unique tracing method for towel spreading using the inchworm grippers is also discussed. Experimental results have demonstrated the effectiveness of both the proposed grippers and also the algorithm. This work was presented in part at the 12th International Symposium on Artificial Life and Robotics, Oita, Japan, January 25–27, 2007     

三维数字服装的研究与探索

采用[TC]2(The Textile Clothing Technology Corporation)扫描仪获得的人体数据,通过对数据进行分析和处理,找到服装造型的各个关键点及关键线,采用原型法生成和真实人体匹配的个性化服装模块,实现了服装上身部件的标准化和模块化生产,最后将模块合并成服装.用户可以根据需要进行调整和修改,提出的方法对三维数字服装的研究和发展提供了新的思路.     

A benchmark for clothes variation in person re-identification

Person re-identification (re-ID) has drawn attention significantly in the computer vision society due to its application and research significance. It aims to retrieve a person of interest across different camera views. However, there are still several factors that hinder the applications of person re-ID. In fact, most common data sets either assume that pedestrians do not change their clothing across different camera views or are taken under constrained environments. Those constraints simplify the person re-ID task and contribute to early development of person re-ID, yet a person has a great possibility to change clothes in real life. To facilitate the research toward conquering those issues, this paper mainly introduces a new benchmark data set for person re-identification. To the best of our knowledge, this data set is currently the most diverse for person re-identification. It contains 107 persons with 9,738 images, captured in 15 indoor/outdoor scenes from September 2019 to December 2019, varying according to viewpoints, lighting, resolutions, human pose, seasons, backgrounds, and clothes especially. We hope that this benchmark data set will encourage further research on person re-identification with clothes variation. Moreover, we also perform extensive analyses on this data set using several state-of-the-art methods. Our dataset is available at https://github.com/nkicsl/NKUP-dataset .     

An approach for capturing the Voice of the Customer based on Virtual Prototyping

The need for companies to improve their competitiveness may lead to innovation and the reconceptualization of traditional products and processes, with companies making an effort to enhance product elements related to functionality, attractiveness, technology and sustainability, and implementing mass-customisation concepts. Mass-customised products are developed to satisfy specific customer needs, in line with increasing demand for product variety and customisation. The analysis of what customers really want, capturing the Voice of the Customer (VOC), is one of the strategies used to establish effective product development processes. Using a VOC survey, it is possible to transform customer needs into the functional and psychological requirements of the product. This paper presents a methodology based on Virtual Reality (VR) technologies to support the capturing of the VOC in regard to the visual, haptic and auditory characteristics of products. This method can be applied to the beginning of the product development process, to allow companies to deduce from the data the requirements of new industrial customised products. A flexible and interactive Virtual Prototype (VP) of a product category is then developed as a product platform in a draft version by designers and configured according to customer needs, using an immersive VR environment. This method, based on the use of VP, reduces the number of physical prototypes that need to be manufactured during the product development process, thus reducing overall costs. In addition, the VP based method supports the mass-customisation process of products through the real-time integration and collection of data for product configuration preferences, involving as many users as possible representative of the target users of the new products. To demonstrate this process a case study concerning the development of the VP for a washing machine, a summary of test sessions with users and results are presented. Specifically, the results presented in this paper are related to improvements in capturing the VOC and reductions in Virtual Prototyping cost and time.     

Women's working wardrobes: a study using card sorts

Abstract: Picture sorts were used to investigate perceptions of women's office clothes, with a sample of ten male and ten female subjects who normally worked in an office environment. The pictures on the cards were taken from catalogues, and showed women's outfits which might be worn in an office. The subjects sorted the cards repeatedly and generated criteria and categories of their own choice. Some of the criteria and categories had not been previously reported in the clothing research literature. Over half of the male subjects, but none of the female subjects, used ‘married/unmarried woman’ as a sorting criterion, although only one of the images sorted showed a wedding ring. A significantly higher proportion of male than of female subjects used dichotomous categorization (i.e. sorting the cards into two piles for one or more of the criteria). The reasons for this are obscure, but do not appear to be a simple outcome of males not knowing much about female clothing. Previous research into clothing has tended to involve researcher‐centred approaches such as semiotics; the results from this study suggest that there would be advantages in wider use of subject‐centred approaches such as card sorts, both in this domain and elsewhere. It was concluded that card sorts were a useful method and should be more widely used.     

Design a personalised product service system utilising a multi-agent system

To help achieve sustainability, reduce the risk of uncertainty, and fulfil a variety of customer needs, the concept of a Product Service System has evolved for use as a business model. While previous studies have addressed PSS development, there is yet no dynamic methodology enabling a PSS to adjust quickly to external changes and customer response. With that gap in mind, this paper presents a Multi-Agent-based Personalised Product Service System (MAPPSS) system. MAPPSS provides a reliable method for analysis which is applicable in various pervasive environments. With this proposed system, users are able to select expected service characteristics, after which service composition is conducted by selecting related products and services from a database. In a comparison of existing systems using different scenarios, we show that customer acceptance, energy utilization, and product utilization generate a two-digit improvement with the proposed system, while profit and costs may increase slightly. When using the proposed method, the competence of an enterprise can be enhanced as a result of these user-oriented features. The performance monitoring and iterating elements of the system allow an almost immediate improvement in the generated PSS.     

基于联合损失胶囊网络的换衣行人重识别

目前的行人重识别（Re-ID）研究主要集中在短时间情形,即一个人的衣着不太可能发生改变的情况。然而现实中更常见的是长时间的情况,这时一个人有很大的机会更换衣服,Re-ID模型应该考虑这种情况。为此,研究了一种基于联合损失胶囊网络的换衣行人重识别方法。所提方法基于换衣行人重识别胶囊网络ReIDCaps,使用与传统的标量神经元相比包含更多信息的矢量胶囊,用其长度表示行人身份信息,用其方向表示行人衣着信息;采用软嵌入注意力（SEA）防止模型过拟合;使用特征稀疏表示（FSR）机制提取具有判别性的特征;增加标签平滑正则化交叉熵损失与Circle Loss的联合损失以提高模型的泛化能力和鲁棒性。在三个换衣行人重识别数据集Celeb-reID、Celeb-reID-light和NKUP上进行实验,实验结果表明所提方法与目前已有的Re-ID方法相比具有一定优势。