Co-Designing Fashion with AI: A Small-Data Approach to Generative Garment Design

This paper details a case study of implementing a co-design framework for integrating artificial intelligence (AI) into fashion design through a collaboration between the researcher and a garment designer. Using the designer’s fashion development as a practical example, this research investigates AI’s potential to enhance creative workflows, improve efficiency, and facilitate marketing strategies that reduce the necessity for physical prototyping. The project implements a small-data approach, exclusively fine-tuning AI models on the designer’s sketches and the researcher’s photography to ensure highly personalised outputs. AI architectures, such as Stable Diffusion and ControlNet, are fine-tuned and used to generate sketches and photorealistic visualisations that work as an extension of the designer’s artistic vision while promoting sustainable awareness and ethical data practices. Through continuous dialogue and a final interview, the designer’s perspective was analysed, demonstrating AI's efficacy as a creativity support tool that streamlines design iteration by enabling rapid prototyping and ideation, while also addressing concerns about preserving artisanal expertise. This research contributes to the growing field of AI in the creative industries by demonstrating the value of co-design, small-data strategies, and ethical AI development to build tools that empower designers.