Generated AI models streamline fashion design with new text and image creation

Generated Artificial Intelligence (AI) has the potential to revolutionize fashion design. By recognizing data patterns and generating new text and images, the AI model is equipped with deep learning algorithms that help fashion designers develop new catalogs, expand creativity, increase efficiency, and deliver products to the market faster.

Large-scale language models (LLMs) such as CHATGPT and AI image generators such as Dall-E have already shown promising results in many industries, and are popularizing the use of AI. In fashion, LLMS can help even designers and non-experts understand past styles and predict future fashion trends. These insights can be used to generate AI image generator prompts to create real fashion collections. As such, it is becoming increasingly important to understand how to effectively integrate AI into fashion.

In a recent study, Professor Yoon Kyung Lee of the Faculty of Clothing and Textile Sciences at Pusan National University in Korea and Professor Choi Hee Liu of the Master’s Department explored how AI can contribute to visualizing seasonal fashion trends.

“To effectively use AI in fashion, we need to understand the characteristics of the generated AI model and make informed judgments about where it can be applied,” explains Professor Lee. “This study studied how to use effective prompt engineering to generate realistic fashion collection images via AI.” Their research was published in The Clothing and Textiles Research Journal on June 22, 2025.

Using CHATGPT-3.5 and CHATGPT-4, the researchers first analyzed fashion trends in men based on historical data up to September 2021. From this analysis, they further used chatGPT to predict fall/winter 2024 men’s fashion trends.

Additionally, Vogue’s design elements in Autumn/Winter 2024 men’s fashion trend data were used as “correction codes” and “codes from literature” from the literature on the concept of fashion design. These elements were then analyzed and reorganized into six final codes: trends, silhouette elements, materials, important items, clothing details and decoration.

I used these code to create 35 prompts for Dall-E 3. Each explains a unique outfit. Prompt followed a consistent template featuring male models walking the runway at the fall/winter 2024 fashion show. This template now allows you to customize a variety of details of actual fashion collection events, including aspect ratio, event, camera angle, model appearance and height, runway design, background, viewer details, and mood. Each prompt was run three times, generating a total of 105 images.

Dall-E 3 was able to fully implement the 67.6% prompt. Specifically, prompts using adjectives showed a high implementation rate. Some images from the generated collection were very similar to the actual fall/winter 2024 boys fashion collection. However, there was an error. Most images leaned towards fashions that you take immediately, and Dall-E struggled to incorporate trending elements like gender fluidity. Trend keywords alone are insufficient to produce accurate results, indicating the need for further learning.

“Our results highlight the important role of fashion experts and show that the implementation of accurate fashion designs for generating AI requires professionally expressed prompts,” Professor Lee said. “With further learning and improvement, generation AI models like the Dall-E 3 will help fashion designers create an entire fashion collection more efficiently while supporting creativity, helping non-experts to understand fashion trends.”

Overall, generative AI has shown that it can be a powerful tool not only for experts but also for the public, making it easier than ever to explore, predict and style seasonal fashion with confidence than ever before.