Methods and systems for hair virtual try-ons (VTOS) using generative ai

A computing system using a diffusion-based Gen AI model with spatial conditioning and 3D alignment for hair simulation addresses the challenge of unpredictable hair style simulations, offering precise and realistic virtual try-on experiences.

WO2026139596A1PCT designated stage Publication Date: 2026-07-02LOREAL SA
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
LOREAL SA
Filing Date
2025-12-23
Publication Date
2026-07-02

AI Technical Summary

Technical Problem

Existing generative artificial intelligence (Gen AI) models struggle to provide precise and photorealistic simulations of desired hair styles, such as new hairstyles or hair coloring, leading to unpredictable results and inadequate user experiences in virtual try-on applications.

Method used

A computing system employing a diffusion-based Gen AI model and a conditioning network, utilizing spatial conditioning with a hairstyle mask and edge detection results from a 3D model, to control the generation of new hairstyles, combined with deep neural networks for hair segmentation and inpainting, to achieve precise and realistic virtual try-on experiences.

Benefits of technology

The system enables users to make informed decisions about hairstyles before application by providing controlled and realistic virtual try-ons, enhancing user experience and ensuring accurate simulation of desired hair styles.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure EP2025088964_02072026_PF_FP_ABST
    Figure EP2025088964_02072026_PF_FP_ABST
Patent Text Reader

Abstract

In embodiments, computer systems and methods provide a hairstyle VTO experience. A generative artificial intelligence (Gen AI) model (e.g. a diffusion-based model and a conditioning network) configured to generate output images in response to spatial conditioning is invoked with a plurality of conditioning images to condition the generation of the new hairstyle on an image of the face. The conditioning images comprise a hairstyle mask to control a shape of the new hairstyle; and an edge detection result image to guide a structure of the new hairstyle. The hairstyle mask and edge detection result are obtained from a 3D model of a sample hairstyle. The mask is aligned with a pose of the face, for example, determined from a face mesh generated for the face. The image of the face is preprocessed to remove an existing hairstyle (e.g. using a dilated hair mask and inpainting).
Need to check novelty before this filing date? Find Prior Art