×

L’Oréal launches AI-powered digital skin diagnostic

Published
today Feb 19, 2019
Reading time
access_time 2 minutes
Share
Download
Download the article
Print
Click here to print
Text size
aA+ aA-

L’Oréal has developed a digital skin diagnostic which uses historic data and selfies to accurately evaluate ageing skin and deliver tailor-made skincare suggestions.



It builds on work carried out by L’Oréal’sResearch & innovation teams over the last 15 years, and uses an artificial intelligence-powered algorithm developed by ModiFace - the augmented reality and artificial intelligence firm it acquired last year.

L’Oréal said the tool uses deep learning to detect patterns based on 6,000 clinical images of men and women in France, China, japan, India and the US; and over 4,500 smartphone selfies, taken by Asian Caucasian and Afro-American in four different lighting conditions.

“With the acquisition of ModiFace, we have started a second phase of L’Oréal’s digital transformation, focused on reinventing the beauty experience through technologies such as voice, AR and AI,” Lubomira Rochet, chief digital officer of L’Oréal, said.

“We believe that services will be the new gateways for discovering our brands and products. After virtual makeup try on, virtual hair color try on, online beauty consultations powered by AR, L’Oréal and ModiFace are proud to unveil the first skin care diagnostic authentically powered by AI and science.”

The first application of the new skin diagnostic launched in the form of Vichy SkinConsult in Canada in January, and will be rolled out on the brand’s websites worldwide later this year.

To use it, women can upload a selfie on the website and get information about their personal skin strengths and priorities to act on based on seven aging variables: under-eye wrinkles, lack of firmness, fine lines, lack of radiance, dark spots, deep wrinkles and pores. Then, each user receives a tailored made product routine to address her specific needs.

Copyright © 2019 FashionNetwork.com All rights reserved.