Towards enhanced creativity in fashion: integrating generative models with hybrid intelligence

Area of Science:

• Artificial Intelligence
• Fashion Technology
• Computational Design

Background:

• Large language models (LLMs) and generative intelligence are increasingly impacting various industries.
• The fashion industry is exploring AI for innovation in design, production, and retail.
• Current LLM applications in fashion face limitations in understanding spatial parameters and design intricacies.

Purpose of the Study:

• To explore the role and potential of LLMs and generative intelligence in the fashion industry.
• To analyze current applications and limitations of LLMs within fashion workflows.
• To propose a hybrid intelligence approach to address identified challenges.

Main Methods:

• Analysis of current LLM applications in fashion.
• Identification of limitations in spatial understanding and design detail processing.
• Proposal of a hybrid intelligence framework for fashion applications.

Main Results:

• LLMs demonstrate significant potential for innovation, personalization, and enhanced customer interaction in fashion.
• Key limitations include the need for improved spatial comprehension and tools for iterative design.
• A hybrid intelligence approach is suggested to overcome current LLM constraints.

Conclusions:

• LLMs offer substantial opportunities for the fashion sector.
• Further development is required to enhance LLMs' capabilities in spatial reasoning and design.
• Future research should focus on hybrid intelligence solutions to fully leverage AI in fashion.