by 
Are we at a point where generative artificial intelligence can predict all fashion trends? Not quite, but an early test showed some results of the eyebrow declaration.
Generative AI models such as Chatgpt and the image generator Dall-E could help improve design efficiency and predict emerging fashion trends, according to a new study by Pusan National University researchers in Busan, South Korea.
More from WWD
While many designers and other creative questions are asked to chat, the generative AI can revolutionize fashion design by, according to researchers, placing patterns in data in data and generating new text and images. They claim that AI models, which are driven by Deep Learning -Algorithms, can help designers develop new catalogs, and expand their creativity with additional efficiency that help to bring products to the market faster.
Professor Yoon Kyung Lee headed the study with Chahi Ryu, a doctoral student of the Pusan National University’s clothing and textiles, to examine how generative AI can contribute to visualization of seasonal fashion trends. Lee said, however, that the results show that “professionally formulated requests are required to implement the fashion design of generative AI”, which emphasizes the important role of fashion experts. Lee has an understanding of the model landscape after working in Milan as a designer for her Uginio brand, which was shown at Seoul Fashion Week and Paris at a time. With more learning and improvements, generative AI models designers will help to create collections more efficiently, she said. This will support their creativity and help “non-experts” to understand fashion trends, added Lee.
Lee and Ryu examined how effective the fast engineering can be when used to generate realistic fashion collection images via AI. Using Chatgpt-3.5 and Chatgpt-4, the researchers first analyzed men’s fashion trends, based on historical data until September 2021. This analysis was then used with Chatgpt to predict men’s fashion trends for autumn 2024. The trends’ design elements predicted as “initial codes” and design elements from Vogue’s 2024 Fashion trend-trend-fashion trend data were used as “modified elements from Vogue’s 2024 men’s fashion trend data. All elements were then analyzed and copied in six codes – trends, silhouette elements, materials, key elements, clothing details and decorations.
Researchers used these six codes to create 35 input requests for Dall-E 3, Openais Image Generator, to describe a unique outfit. The input requests were applied to a template in which a male model went along a runway at a fashion show 2024. For the additional effect, the design of the runway, the background, the camera hob, the appearances and heights of the models and the size of the public details could be set. The researchers carried out each command prompt three times and generated a total of 105 pictures.
According to the study, Dall-E 3 in 67.6 percent of the cases was “perfectly implementing the input requests”. (Entry requests with adjectives are said to have shown a high implementation rate.) According to researchers, some of the images in the generated collection were “very similar” to the actual men’s fashion collections of 2024 men. However, there were some mistakes in which most images looked ready as possible, and according to the study, Dall-E fought to integrate trend elements such as gender fluidity. Lee and Ryu found that trend keywords alone were not enough to achieve precise results, and further learning is required.
However, the implementation of AI in fashion requires understanding the properties of generative AI models and “well -founded” decisions about where they can be applied, the study says.
After the fashion designer Norma Kamali Ai studied at the Massachusetts Institute of Technology, he said on Friday: “AI is great in using data and statistics to create algorithms that fit well into trends.” She added: “The trends differ greatly from set a trend, This is an unexpected event that creates a movement in a completely different direction. “
Kamali suggested that AI data that show an increase in obstetricians could potentially lead to new trends in maternity use for Gen -Z mothers. “However, original ideas can best be created by a person with passion, soul and feelings,” said Kamali.
Your former professor Abel Sanchez, the Executive Director of the Geospatial Data Center, found: “When it comes to AI in general, we try to predict the future with the past. When it comes to fashion, we can predict many things, but not benefit. Other things that may take place.”
He compared Ki with the way Renaissance artists like Leonardo da Vinci worked with trainees, and at a certain point in time they directed the work and did not form them with their hands. “It is still your vision that is created, but they don’t move them directly,” he said.
One of the possibilities that you have through the technology is to “navigate through the world in a simpler and more personal way,” said Sanchez. If he realizes that the average person reads 750 books in their life and how many books are around 130 million, he said: “If you calculate this as a percentage, it is a number that is almost zero. The world demands more of us and changes faster and changes faster and does not change. We may be able to adapt. We may not adapt. There are many things that we can no longer have.”
According to Sanchez, there are some things that AI can help people at the moment to overwhelm them at the moment: “One of the possibilities is a easier life. In contrast to the fight against telecommunications [company] Or which institution that has outdated business models and customer service offers something that integrates seamlessly into your life. The transformation will be large and painful. And it will touch all areas of society. I work on an academic facility and we are disturbed in one of the greatest ways. Are we changing soon? We are not. “
When it comes to fashion and the principles that are at work, it is no different, said Sanchez. “Our ability to innovate, explore and create becomes easier.”
While some are of the opinion that trend forecasts requires a certain tendency, Sanchez questioned the idea that there is a little magic and uniqueness for each of us. A typical example, he said, take into account the movement movement, which has led to millions of compromised passwords.
Sanchez recognized how some companies use digital avatars to understand how people react to products and use other digital versions of models in advertising. “One of the challenges is that certain industries are more resistant and creative communities are one of the most constant changes. If I think about writers, the film industry, although this changes, there is a sense that AI is the enemy.”
He pointed out that art did not disappear after the beginning of photography, and said that in fashion and AI: “This technology is here. We have to overcome many of the prefabricated ideas and prejudices against technology. It will take a little longer. But part of it will be a question of innovation and those who enable these new artificials that use these technologies.
Best of WWD
Register for the WWD newsletter. For the latest news, follow us on Facebook, Twitter and Instagram.
