Myeongjin Go, Beauty Tech Lab We have won the CES Innovation Awards for nine technologies over six years, from 2020 to this year. We didn’t initially start development with CES in mind. Wanna-Beauty AI, in particular, began in 2022 with a simple idea: using AI to answer the common customer question, ‘What makeup looks best on me?’ We started joint research with KAIST to accurately analyze facial features, recommend optimal makeup, and generate realistic images with ideal solutions for any situation. By April 2024, our joint research showed visible results, leading to our entry into CES 2025. I see the CES Innovation Award as part of our ‘journey of innovation for New Beauty.’ As an Amorepacific researcher, I take pride in two things: customer-centricity is always our starting point for development, and we don’t stop at technological demonstrations but achieve commercial product launches, just like how our first Innovation Award winner in 2020, the ‘Tailored Facial Mask Pack 3D Printing System,’ was commercialized through IOPE.

Myeongjin Go, Beauty Tech Lab Wanna-Beauty AI analyzes skin tone, facial proportions, and features by mapping 68 points on a user’s uploaded photo and provides makeup recommendations and virtual makeup experiences based on data learned from makeup experts’ expertise. Users can conveniently consult while trying different makeup looks on their faces through a voice chatbot interface. We implemented voice recognition for consumer convenience and to quickly understand preferences through AI conversation and suggest optimal makeup solutions. Through our advanced model developed jointly with KAIST, we focused on providing virtual makeup that looks natural and realistic on each customer’s face.

Myeongjin Go, Beauty Tech Lab To be more technical, it’s a solution that combines Amorepacific’s image diagnosis technology (AI facial segmentation), an LLM (Large Language Model) trained on makeup trends, and Stable Diffusion technology jointly developed with KAIST.
Our facial analysis technique requires high accuracy in a short time because it needs to quickly identify facial features and determine makeup colors and styles that suit the customer—specifically, what to conceal and what to emphasize. If facial analysis is inaccurate, it could lead to inappropriate makeup suggestions, ultimately causing customers to abandon the service. That’s why we conducted in-depth workshops with HERA’s makeup professionals team for facial analysis and used the results to train the AI on the faces of 1,000 Korean women. As a result, we achieved 83% accuracy compared to expert analysis in swiftly diagnosing facial features like eyes, nose, and mouth, including their length, area, and proportions. We plan to upgrade our AI model to achieve even higher accuracy as we advance.

Generating realistic makeup images was another key focus area, where we utilized Stable Diffusion, an image generation AI widely used in models like Midjourney. Since this technique typically generates new images by randomly modifying original images, we worked diligently to overcome this limitation. As a result, we developed technology that maintains the original image while properly implementing desired colors on facial features, enabling us to generate high-quality, ideal virtual makeup.

Myeongjin Go, Beauty Tech Lab We still have a long way to go with this aspect. To truly understand customers’ faces and incorporate our makeup expertise and recent trends, we need extensive datasets and must undergo the process of training AI with professional guidelines. We’re working on more training and professional verification while trying to make good use of LLM. We’re entering a world where there’s nothing AI can’t do.

Myeongjin Go, Beauty Tech Lab In researching and developing new technology, the most important thing is communicating and aligning with people who share the same goal of enhancing customer experience. I knew nothing about AI, but I began seeing its potential thanks to Sungmin Jung from Beauty Tech Lab, who started working on this project. He always proposed solutions whenever we faced challenges. Similarly, Heechan Jeon from the Customization Business Development Team helped us establish guidelines based on comprehensive makeup understanding and experience. I’d also like to thank Jinsoo Lee, Changhoon Baek, Jungki Lim, Minsu Shin, Areum Jo, Minkyung Cha from the HERA Makeup Professionals Team, and Moonha Kim from the Customization Business Development Team. Finally, we couldn’t have come this far without Team Leader Jin Nam, who provided full support and opportunities for entirely new research, and Wonseok Park, who offered technical direction and research support.

Myeongjin Go, Beauty Tech Lab The most crucial source is ultimately ‘customers and colleagues.’ We gained the inspiration and momentum to start Wanna-Beauty AI because we had colleagues who listened directly to what customers really wanted, thought about solutions together, and developed products and services collaboratively.
Moreover, since the CES Innovation Awards attract around 4,000 entries annually, presenting our technology’s appealing points concisely yet impactfully is crucial. Three key evaluation criteria are critical when assessing submitted products or technologies:
1. Contribution to Humanity: Does it benefit individuals and society significantly?
2. Technological Innovation: How does it overcome existing limitations, and how original is it?
3. Commercialization: Has it achieved commercial success and reached many people? Does it have the potential to disrupt existing markets or create new ones?
The secret to winning lies in our ability to detail the innovation and originality of our products and technologies in technical documentation while presenting how customers can easily use the technology and its benefits through image presentations.

Myeongjin Go, Beauty Tech Lab We’re collaborating with the Digital Strategy Unit to commercialize Wanna-Beauty AI. We’d like to review and improve the model to launch it within this year so customers can experience Wanna-Beauty AI.

Myeongjin Go, Beauty Tech Lab From the start of our Wanna-Beauty AI research, our goal has been to help customers discover their ideal beauty—even aspects they weren’t aware of—by accurately interpreting their skin concerns. We believe that as this experience becomes more widespread and accessible to more customers, each individual will be able to discover their unique beauty that they hadn’t recognized before.

Goeun Lee, Beauty Device Team Ta-da~! Here’s a spoiler through this behind-the-scenes talk: makeon’s Skin Light Therapy 3S, which amazed CES 2025, will be launched soon! The Skin Light Therapy 3S will officially launch in March with a dedicated application featuring AI skin analysis and care solutions. The technologies showcased at CES won’t remain distant concepts but become tools customers can use comfortably at home to discover their skin’s natural radiance through makeon.

Hakgwon Park, AI Solution Team The core of our ‘AI Skin Analysis and Care Solution’ is to help customers analyze their skin conveniently and quickly. Customers using this solution can receive rapid skin analysis anytime, anywhere, helping them discover and pursue their unique beauty. Additionally, it provides skin change graphs to help better understand their skin condition and make choosing suitable products easier. We expect this will help realize customers’ true New Beauty.

Myeongjin Go, Beauty Tech Lab While makeup solutions that customers could experience were previously product-centered, the attempt to help customers find their beauty through digital and AI-suggested makeup feels like a New Beauty moment to me. Customer interest and concerns about makeup have always been constant, and at CES, along with our Wanna-Beauty AI, various global beauty companies showcased different solutions in the digital realm. Though we still have a long way to go, I hope these challenges signal the beginning of solutions toward New Beauty.

Hakgwon Park, AI Solution Team I was particularly impressed by a service that analyzes skin protein structure using AI to detect signs of skin aging and provide solutions. While it seems far from commercialization, it made me realize that the era of New Beauty has truly arrived.

Goeun Lee, Beauty Device Team The New Beauty I mainly felt at CES 2025 is Amorepacific’s unique capability expressed through accumulated beauty expertise, customer-centricity, and skin data. Since product technology differentiation is limited, we need a comprehensive approach beyond technical differentiation to express individual beauty. I believe Amorepacific’s clear commitment to providing more benefits to customers is itself the practice of New Beauty.

Myeongjin Go, Beauty Tech Lab As one of Amorepacific’s researchers, I’m deeply committed to developing innovative technologies that provide new experiences and meet future customer needs. We aim to create unprecedented solutions that can become powerful market-leading technologies. We will continue striving to realize New Beauty and technological innovation that customers can genuinely experience through products that amaze the world.

Goeun Lee, Beauty Device Team The successful launch of Skin Light Therapy 3S is writing a new chapter in the beauty device market history. As this market grows, we’ll work hard to satisfy more customers through new devices and services that makeon will launch.

Hakgwon Park, AI Solution Team The beauty tech platform is in the present tense. Beyond CES, we’re preparing for continuous service launches, including Amorepacific brand’s AI care services and global expansion, and we’re developing additional digital AI beauty features besides non-face-to-face skin diagnosis for the first half of 2025. As for non-face-to-face skin diagnosis for international markets, we’re continuously advancing image learning from diverse global customers and sophisticating measurement criteria. We’re working diligently to enable customers to experience more of Amorepacific’s AI-based beauty tech platforms, including services where customers can experience various beauty technologies.