DesignBridge: Bridging Designer Expertise and User Preferences through AI-Enhanced Co-Design for Fashion Conference: 31st International Conference on Intelligent User Interfaces; March 23–26, 2026; Paphos, Cyprus 31st International Conference on Intelligent User Interfaces (IUI ’26), March 23–26, 20

Yuheng Shao, Yuansong Xu, Yifan Jin, Shuhao Zhang, Wenxin Gu, Quan Li · 2026

Read DesignBridge: Bridging Designer Expertise and User Preferences through AI-Enhanced Co-Design for Fashion Conference: 31st International Conference on Intelligent User Interfaces; March 23–26, 2026; Paphos, Cyprus 31st International Conference on Intelligent User Interfaces (IUI ’26), March 23–26, 20 by Yuheng Shao, Yuansong Xu, Yifan Jin, Shuhao Zhang, Wenxin Gu, Quan Li (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-14639.txt.

Introduction

Yuheng Shao and five co-authors build a co-design tool for fashion in this 2026 paper, set for the Intelligent User Interfaces conference. Effective collaboration between designers and users matters for fashion design. It can increase user acceptance of fashion products and thereby create value. The authors' system is DesignBridge, named in their title.

DesignBridge bridges designer expertise and user preferences through AI-enhanced co-design. In co-design, users shape the product alongside the professional. The abstract in the evidence pack cuts off early in that setup. Readers get a system aimed at fashion work where expert craft and wearer taste must meet in one draft.

Catalog record

CREATORYuheng Shao, Yuansong Xu, Yifan Jin, Shuhao Zhang, Wenxin Gu, Quan Li
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE103 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-14639

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

DesignBridge: Bridging Designer Expertise and User Preferences through AI-Enhanced Co-Design for Fashion Conference: 31st International Conference on Intelligent User Interfaces; March 23–26, 2026; Paphos, Cyprus 31st International Conference on Intelligent User Interfaces (IUI ’26), March 23–26, 2026, Paphos, Cyprus DOI: 10.1145/3742413.3789081 ISBN: 979-8-4007-1984-4/2026/03 CCS: Human-centered computing Interactive systems and tools Yuheng Shao, Yuansong Xu, Yifan Jin, Shuhao Zhang, Wenxin Gu, Quan Li Abstract . Effective collaboration between designers and users is important for fashion design, which can increase the user acceptance of fashion products and thereby create value. However, it remains an enduring challenge, as traditional designer-centric approaches restrict meaningful user participation, while user-driven methods demand design proficiency, often marginalizing professional creative judgment. Current co-design practices, including workshops and AI-assisted frameworks, struggle with low user engagement, inefficient preference collection, and difficulties in balancing user feedback with design considerations. To address these challenges, we conducted a formative study with designers and users experienced in co-design (N=7), identifying critical challenges for current collaboration between designers and users in the co-design process, and their requirements. Informed by these insights, we introduce DesignBridge , a multi-platform AI-enhanced interactive system that bridges designer expertise and user preferences through three stages: (1) Initial Design Framing , where designers define initial concepts. (2) Preference Expression Collection , where users intuitively articulate preferences via interactive tools. (3) Preference-Integrated Design , where designers use AI-assisted analytics to integrate feedback into cohesive designs. A user study demonstrates that DesignBridge significantly enhances user preference collection and analysis, enabling designers to integrate diverse preferences with professional expertise. Figure 1. An example scenario illustrating how designers and users can effectively engage in a co-design process using DesignBridge : (A) Designers and users face a gap in collaboration during co-design due to domain differences. (B) In fashion design, users often struggle to accurately and efficiently express their ideas, while designers must balance varying user preferences and navigate challenges such as trade-offs between user preferences and design concepts. (C) DesignBridge facilitates the process by assisting users in expressing their preferences and helping designers analyze user feedback for design integration. (D) This fosters collaboration between users and designers in the co-design process, enhancing both user experience and design outcomes. † † cc-license: by 1. INTRODUCTION In design practice, professionals create products to meet user needs, while stakeholders act as both recipients and contributors through ongoing information exchange ( 75 ) . Designers typically elicit user requirements and preferences through analytical methods, surveys, and interviews to inform solution development ( 5 ) . In fashion design, this process often follows an iterative Design-Display-Evaluation-Adjustment cycle ( 36 ) , emphasizing coherent thematic concepts and user-centered scenarios. Following the establishment of an aesthetic direction and conceptual framework, designers explore multiple potential variations while considering user preferences. Unlike the production-preparation stage, where user involvement focuses on fit, sizing, and manufacturability ( 49 ; 81 ; 48 ) , this stage highlights the value of user collaboration during conceptual exploration and variation generation, thereby enhancing both relevance and creativity. Traditional approaches, however, remain predominantly design-centric, limiting direct user participation and restricting the articulation and integration of user needs within design frameworks. Although user-driven methodologies such as mass customization ( 24 ) and open design ( 11 ; 4 ) empower users to steer design directions, they often require users to possess design expertise and may undermine the role of professional designers, risking diminished control over outputs and misalignment with predefined design concepts. Existing co-design frameworks ( 69 ; 83 ) foster collaborative innovation. In fashion, research has explored co-design to augment designer creativity through human-AI collaboration ( 42 ; 88 ; 45 ) . However, these studies focus primarily on designer-AI collaboration rather than direct user-designer interaction. Alternative modalities, such as workshops and collaborative creation, enable sustained interaction to align outputs with user needs and design concepts ( 86 ; 71 ) . Nevertheless, these methods introduce operational complexities: extended timelines, communication inefficiencies, and difficulties reconciling non-expert input with professional workflows. Consequently, fashion design needs a more efficient, structured interaction paradigm to enhance user-designer collaboration. To understand designer-user co-design challenges, we conducted a formative study with seven designers and experienced participants. Findings show that current processes struggle with effective preference collection and analysis. Low engagement and a lack of fine-grained methods result in incomplete preference capture, while the absence of mechanisms to balance user feedback with design considerations complicates the workflow. These insights inspire us to develop intuitive interactions that enhance user participation and enable designers to effectively fuse diverse feedback with their design intent. Based on the insights from the formative study, we propose DesignBridge , a multi-platform interactive system that supports the co-design process in the fashion design domain. The system has two interfaces: one for users and one for designers. The entire workflow is divided into three stages: Initial Design Framing , Preference Expression Collection , and Preference-Integrated Design . In Stage One: Initial Design Framing , designers input initial design considerations such as the design scene, clothing type, and principles. The system generates a scene image based on the input and filters relevant attributes from the design space to build the design image database. In Stage Two: Preference Expression Collection , users review provided garment designs on a virtual model in the given scene and a standalone garment image. Through brushing and commenting, users express their preferences (like or dislike) on specific parts or the overall design. Each user goes through multiple rounds of interaction to capture their overall preferences. In Stage Three: Preference-Integrated Design , designers examine the collected user preference information. Through a puzzle-like approach, they select attributes from each design space dimension to form a comprehensive design description based on the design space. For each attribute, designers navigate a hierarchical tree structure to examine the associated garments and user interactions. Through pruning, designers decide which attribute-related image data should be included in the generative model for fine-tuning. They can also review the model’s predictions of individual user preferences for the current design, along with the model’s interpretability results across different design space dimensions as a reference. To evaluate DesignBridge , we conduct a technical evaluation and a user study. The technical evaluation uses ratings to assess the effectiveness of the two fine-tuning processes in improving Design Space Capture and User Preference Integration, compared to the outputs generated by VisionRealistic v2 FluxDev 1 1 1 https://civitai.com/models/619656/vision-realistic and GPT-4o 2 2 2 https://chatgpt.com . The user study further evaluates DesignBridge ’s effectiveness and usability from both the user and designer’s perspective in the co-design process. The results indicate that the system could facilitate clear and efficient preference expression by users, as well as effective analysis of user feedback and integration of design considerations by designers during the design generation process. In summary, our contribution can be summarized as follows: • We conduct a formative study to identify the challenges and needs in the interactions between designer and users in the co-design process. • We propose DesignBridge , an interactive multi-interface system that supports the co-design process between designers and users. • We conduct a technical evaluation and user study to collect the feedback from designers and users, evaluate the effectiveness and usability of DesignBridge . 2.

Read the complete text