South Korea’s Lululab leads the market with AI-driven skin-deep solutions
Lululab, a South Korean startup and spin-off company from Samsung Electronics, is being called the top runner in AI-driven skin analysis and cosmetics product matching technology. This is because they handle so many aspects of their business in-house; from the devices for skin data collection and algorithms for analysis to cosmetics data collection and product matching algorithms.
Lululab is a South Korean startup that is a leader in skin analysis and cosmetics product matching. Specifically, they analyze facial image data with AI and make personalized cosmetics recommendations.
Lululab is one of the spin-off companies from Samsung Electronics’ in-house venture program C-Lab. Currently, they’re developing and deploying the AI skin diagnosis product LUMINI, which can perform the whole process of analyzing skin, making a diagnosis and recommending ideal cosmetics, all in the short space of around 10 seconds.
This startup has won the Innovation Award two years in a row (2019 and 2020) at the Consumer Electronics Show held in Las Vegas, the world’s largest electronic appliance trade show. Previously they also participated in the LVMH startup support program Luxury Lab.
Founder and CEO Yongjoon Choe majored in Bio-Engineering at Cornell University in the US before entering Samsung Electronics that, at the time, had declared a shift in business from smartphones towards health care. During this time, Choe thought about what proposal he could submit to C-Lab and came up with the idea that skin data could become an indicator of people’s health. He also realized that if he could concoct a way to use skin data to predict the likelihood of someone getting any disease, he could bring about innovation in the health care field.
However, Choe found that “at the time, when I approached hospitals, although I thought they would naturally have skin data, there was no such data to be found.” Therefore, he decided to first propose a way of analyzing images of faces and managing skin data. This led to the development of LUMINI.
His choosing to enter the beauty market was because beauty users were more likely to adopt new products — in other words, it was easier to collect skin data in the beauty industry. Choe thought that “even if you’ve got the technology, there’s no point if it’s not used”, and he aimed to narrow down his target user to beauty consumers who want to more efficiently and precisely know their own skin condition and what products will allow them to improve their skin.
LUMINI is basically made up of three elements: a device (hardware) that acquires the skin data, an AI algorithm (software) that analyzes that data, and a data managing system that visualizes the acquired data onscreen.
LUMINI’s strength lies in the accuracy of its data collecting, that involves scanning the condition of the skin down to its deeper layers, and its advanced powers of analysis.
Its accuracy allows for results equivalent to diagnoses clinical experts perform with the naked eye. Along with this, LUMINI also excels at providing consistent analysis results even under varying environmental factors, such as the light brightness, temperature or humidity in a room.
There are other types of skin analysis systems already, such as those that use selfies taken with a smartphone. However, most adopt a method of using only one part of the skin for measurement and aren’t able to easily obtain consistent results under varying conditions when analyzing someone’s overall skin. LUMINI, using its custom internally-developed scanner, not only brings about highly accurate results but also has high levels of “reproducibility” (90% on average).
Also, in the initial stages of developing LUMINI, the standard healthy skin conditions for referencing during analysis were jointly developed with medical specialists from Seoul National University. Furthermore, these standard values are being continuously supplemented using the data obtained from the products released in each country.
Choe boasts that “currently there are mostly no beauty tech solutions based on facial recognition AI that are at the same level as us.”
So far, Lululab has developed the “LUMINI Kiosk”, a standalone small-sized information terminal, and “LUMINI”, a version for professional consultants.
The kiosk type can be installed even in small-scale spaces and is made for use in retailers, including department stores. For instance, it can be placed at the entrance to cosmetics-selling areas in malls as a way for users to analyze their skin data themselves. Users can have their recommended products listed up in the kiosk and be guided to the appropriate stores.
Choe reveals their current state of usage: “The kiosk type that analyzes users’ skin and recommends cosmetics sold in the vicinity can itself be designated as a “beauty AI store”. In one currently operating case, the AI store gets around 4,000 customers on average per month, and around half of them visit the cosmetics stores and additionally talk to the staff there based on the analysis results.”
On the other hand, the professional consultant version of LUMINI is a small-sized solution that utilizes a compact and portable device. It’s useful for stores that perform one-on-one counseling, such as beauty salons or beauty dermatologists.
On top of this, Lululab is planning this summer to introduce to the market LUMINI HOME, a tabletop smart mirror version of the product for general consumers.
Choe enthusiastically reveals that future-wise “we want to create solutions that, through skin data, can also make recommendations for people’s health and lifestyles.”
Text: Ching Li Tor
Original text (Japanese): Jonggi HA
