Application of a Product Styling AI Diagnosis System in Industrial Design Education



Abstract Book of the 10th World Conference on Teaching and Education

Year: 2026

[PDF]

Application of a Product Styling AI Diagnosis System in Industrial Design Education

Prof. Dr. Yung-Hsun Cheng, Yuan-Chun Cheng

ABSTRACT:

With the advancement of artificial intelligence and digital transformation, industrial design education increasingly needs methods that connect visual judgment, computational thinking, and evidence-based design discussion. Traditional product styling courses depend heavily on precedent analysis, visual comparison, and instructor critique; although these approaches remain essential, they often provide limited quantitative evidence for explaining how product forms communicate meanings such as minimal, expressive, functional, streamlined, traditional, or high-tech. This study presents a product styling AI diagnosis system for industrial design education. Plumbing hardware products were selected as teaching cases because they show clear variations in form, proportion, line, surface, material expression, and market positioning. The system integrates image data collection, style semantic annotation, CLIP-based image-text similarity analysis, and PCA/t-SNE visualization. A database of more than 1,000 product images was organized, and a vocabulary framework was built around four style axes and eight categories. The CLIP model calculated semantic similarity between product images and textual descriptors, while visualization methods generated style maps for interpreting clusters and positioning. Expert feedback from 16 plumbing hardware-related company representatives indicated strong recognition of the system’s usability, style identification capability, and educational value. The findings suggest that AI-based image analysis can transform intuitive style perception into explicit, comparable, and discussable information, supporting product analysis, critique, market positioning, and concept development within a human-AI collaborative learning model.

Keywords: Industrial Design Education; Product Styling Diagnosis; Artificial Intelligence; Design Pedagogy; Human-AI Collaboration