Cosmetic Transformation Through Image Synthesis

Number: 20200051301
Published: 2020-02-13
Filed: 2019-10-22
Assignee: Disney Enterprises, Inc.
Inventors: Fidaleo; Douglas A.
CPC: G06T11/60; H04N9/3185
Verdict: Low Notable software
Source: Google Patents · FreePatentsOnline

The keeper's note

ML image-synthesis cosmetic transformation technique.

Abstract

According to one implementation, a system for performing cosmetic transformation through image synthesis includes a computing platform having a hardware processor and a system memory storing an image synthesis software code. The hardware processor executes the image synthesis software code to receive a user image depicting a user of the system, to receive a reference image for transforming to resemble the user image, the reference image projecting an identifiable persona, and to generate a user image metadata describing the user image, based on the user image. The hardware processor further executes the image synthesis software code to identify to a reference image metadata describing the reference image, and to transform the reference image to resemble the user image, based on the user image metadata and the reference image metadata, while preserving the identifiable persona projected by the reference image.

Background

BACKGROUND

In psychology, mirroring refers to human interaction in which one person imitates the gestures, vocal tone, or attitude of another. Mirroring typically occurs subconsciously and can result in the establishment or deepening of rapport between the person performing the mirroring and the person being mirrored. That is to say, mirroring can create an environment in which a person tends to find another person more interesting and to experience their company as more enjoyable than might be the case in the absence of mirroring behavior.

Due to its basis in human-to-human interaction, psychological mirroring does not spontaneously occur in interactions between people and virtual or dramatic characters, or between people and physical objects. As a result, the increasing diversity in appearance, physiognomy, decoration, and dress reflected by many virtual characters encountered in theme park attractions, games, and movies, for example, as well as the increasing cultural diversity of the general public, may make it difficult for some members of the public to identify with and enjoy those virtual characters.SUMMARY

There are provided systems and methods for performing cosmetic transformation through image synthesis, substantially as shown in and/or described in connection with at least one of the figures, and as set forth more completely in the claims.

Claims

21. A cosmetic transformation system comprising: a computing platform including a hardware processor and a system memory; an image synthesis software code stored in the system memory; the hardware processor configured to execute the image synthesis software code to: receive a user facial image depicting a user of the cosmetic transformation system; receive a reference facial image for transforming to resemble the user facial image, the reference facial image projecting an identifiable persona; obtain first metadata describing first key geometric features of the user facial image; obtain second metadata describing second key geometric features of the reference facial image; and transform the reference facial image to resemble the user facial image, based on the first metadata describing first key geometric features of the user facial image and the second metadata describing second key geometric features of the reference facial image; wherein transforming the reference facial image preserves the identifiable persona projected by the reference facial image. 31. A method for use by a cosmetic transformation system including a computing platform having a hardware processor and a system memory storing an image synthesis software code, the method comprising: receiving, using the hardware processor, a user facial image depicting a user of the cosmetic transformation system; receiving, using the hardware processor, a reference facial image for transforming to resemble the user facial image, the reference facial image projecting an identifiable persona; obtaining, using the hardware processor, first metadata describing first key geometric features of the user facial image; obtaining, using the hardware processor, second metadata describing second key geometric features of the reference facial image; and transforming, using the hardware processor, the reference facial image to resemble the user facial image, based on the first metadata describing first key geometric features of the user facial image and the second metadata describing second key geometric features of the reference facial image; wherein transforming the reference facial image preserves the identifiable persona projected by the reference facial image.