Open access

An Ai-Driven Framework for Personalized Fashion Design Generation and Customization

Madilyn A. Marquez Nolasco Malabago Gregorio P. Pajaron Janice A. Alivio J. Garcia Mario P. Rebucas M. E. Medalla

2026 · International journal of research and scientific innovation · 0 citations

Abstract

Artificial intelligence (AI) is increasingly used to accelerate ideation and visualization in fashion, yet empirical evidence on localized, education-oriented design systems remains limited. This study developed and evaluated Fashion Studio, an AI-driven framework that converts user prompts, sketches, style preferences, fabric selections, and body-related parameters into customizable fashion-design concepts. The framework was developed through a system development life-cycle process and organized around input, AI processing, customization, visualization, project management, and designer-client feedback modules. A descriptive-developmental evaluation was conducted at Palompon Institute of Technology during School Year 2025-2026. Thirty students and 10 faculty members (N = 40) assessed the system using a structured questionnaire adapted from ISO/IEC 25010 quality characteristics. Descriptive statistics summarized functional requirements, user familiarity, software acceptability, and open-ended feedback; group evaluations were compared at the .05 significance level. The user interface (M = 4.50) and design-generation process (M = 4.47) were rated Excellent, whereas customization (M = 4.39) was Very Good. Overall acceptability was Excellent (grand M = 4.42). User satisfaction received the highest criterion score (M = 4.56), followed by functionality (M = 4.48) and usability (M = 4.45). Reliability (M = 4.28) and efficiency (M = 4.30) were comparatively lower but remained Very Good. Student (M = 4.49) and faculty (M = 4.37) evaluations did not differ significantly (p = .178). Respondents primarily recommended faster loading, a broader template library, improved body-measurement modeling, and image-based styling input. The findings indicate that a human-centered AI platform can support rapid ideation and collaborative design in fashion education, although technical optimization and more granular personalization are necessary before wider institutional deployment.

DOI Semantic Scholar

Read PDF