How generative AI is changing the fashion industry

Opé's journey

Opé started her experimentation by using OpenAI's generative AI image creation system, Dall-E, and later switched to another generative AI imaging platform, Midjourney.

As she became confident in her AI-generated designs, she entered the AI Fashion Week competition in New York in April and partnered with design clothing brand Revolve. She worked with Revolve to edit her AI-generated designs and ended up placing third. Her designs sold out on the Revolve website right away, she said.

"I didn't expect to win," she told TechTarget Editorial. "When I realized that AI was getting better and better for fashion, I was like, 'I can really use this to transform how I'm communicating in pictures, because that was always important to me as a fashion stylist.'"

While she was a finalist in the competition and has been gaining success with her designs, Opé also has faced criticism from some who shun her use of AI to create designs.

While many other artists have found success with generative tools like Dall-E and Midjourney, others see the tools as an intrusion into their creativity. Some artists have accused the vendors of the generative AI technologies of stealing artists' work.

However, Opé does not view AI as replacing her job or the jobs of fashion designers but rather as an assistant.

"AI is a servant," she said. "It's a helper, and it can include more people in the process. It can help us communicate faster."

While there have long been applications for AI in fashion, generative AI is expected to add $150 billion in revenue to the industry over the next three to five years, according to McKinsey & Company.

Virtual try-on

Generative AI is not only changing fashion design but also skin care. An interactive virtual try-on booth at the conference illustrated this.

Created by Perfect Corp., a vendor of AI and augmented reality (AR) for skincare, fashion and beauty, the virtual try-on technology uses AI and AR -- supplementing a user's real environment with digital data -- to help users analyze their skin or try different beauty products.

Perfect works with companies such as Estée Lauder, which uses Perfect's virtual makeup try-on system on its website to let customers try on different products.

Perfect's AgileFace technology is a face-tracking system that creates a precise analysis of a customer's face and ensures the products they try on are depicted realistically.

The company's Virtual Styler lets users experiment with different hairstyles. For Virtual Styler, Perfect uses the image-generating AI platform Stable Diffusion. Perfect trains the AI engine to create different hairstyles that fit different face types.

"For example, we give it some prompts, likem 'Create a Type 2 style,'" said Wayne Liu, the company's chief growth officer and president of the Americas. Users give the generative AI engine prompts about hairstyle and length so the system can learn how the hair looks and then try to mimic and create that style.

Another beauty company using the Virtual Styler for consumers is Tresemmé.

Melissa Noorily, a product manager for JP Morgan Chase, was among the conference attendees captivated by Perfect's technology. While Noorily is not in the beauty industry, she recently took an interest in skin care.

"I think it's super nice to have it scan your face and then give you numbers and readings on the different aspects of your skin and where they are," she said.

Revea's technology

Another skincare company using AI technology to improve beauty is Revea.

Revea has consumers take a skin assessment quiz using their phone. Its patented face scan technology scans and analyzes the user's face. The company then uses the results to create custom formulations for customers.

AI is a servant. It's a helper, and it can include more people in the process. It can help us communicate faster.

Revea's goal was to not only capture and analyze skin types using computer vision but also to go behind the camera and under the skin to figure out what might be causing skin problems, founder and CEO Chaz Giles said during a presentation.

Revea uses some of the technology used in biology, such as spatial frequency domain imaging -- which projects a 2D pattern of light and analyzes it -- along with learning models, neural networks and special imaging cameras. With those tools, Revea determines the deep biological structure underneath a person's skin and produces a formulation to improve their skin.

"We have all the dermatological measurements," Giles said in an interview. "That's how we train our models, and that's how we check the accuracy and validation."

As an expert in technology and fashion, Giles thinks AI will continue to change the industry.

"You're seeing a lot in GenAI and visualizations and text," he said. "The better those models get, the better communication can be tailored to an individual."

Esther Ajao is a TechTarget Editorial news writer covering artificial intelligence software and systems.