3D Clothing Display with Face Replacement for Virtual Try-On

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing virtual try-on systems struggle to accurately represent how clothes would fit a user, especially when the mannequin's face does not resemble the user's, leading to difficulties in determining fit and potentially eerie impressions due to realistic imagery.

Innovation Solution

A display system that generates and processes three-dimensional data representing a mannequin wearing clothes, allowing users to specify desired appearances, including face replacement, blurring, and background combinations, to create a personalized and realistic virtual try-on experience.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a mannequin with a face significantly different from the user's is used, then the body shape fit can be accurately displayed, but the user has difficulty determining whether the clothes become the user

Engineering Contradiction:
Improvebody shape fit accuracyVSAvoiduser identification with mannequin
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent applies copying by replacing the mannequin's face with the user's actual face image. This creates a composite image where the user's face is copied onto the mannequin's body, allowing users to easily identify with the display while maintaining accurate body shape representation. The face replacement technique directly resolves the contradiction by making the mannequin resemble the user without altering the body shape data.

Inventive Principle:
Principle #26Copying

2Ease of operation

If a user's face picture is combined with the body of a mannequin to make it look like the real user, then user identification is improved, but the image creates an eerie impression

Engineering Contradiction:
Improveuser identification with mannequinVSAvoideerie impression
Core Design Contradiction:
Ease of operationVSObject-affected harmful factors

Solution Approach 1:

The patent applies local quality by selectively applying different processing to different parts of the mannequin. The face region receives special treatment through blurring or replacement techniques, while the body maintains realistic proportions and features. This localized differentiation allows the face to appear more natural and less eerie, while preserving the overall realistic appearance needed for accurate clothing display.

Inventive Principle:
Principle #3Local quality

3Reliability

If realistic imagery is used to show the user wearing clothes, then the display is more convincing, but it creates an eerie impression when the mannequin looks like a real human

Engineering Contradiction:
Improvedisplay convincingnessVSAvoideerie impression
Core Design Contradiction:
ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The patent uses copying to replace the mannequin's face with the user's actual face image, creating a composite that maintains realistic body proportions while featuring the user's actual facial characteristics. This approach preserves the convincingness of the display by showing the user's real face rather than a generic or distorted one, while avoiding the eerie effect by using authentic user features instead of attempting to create a hyper-realistic mannequin.

Inventive Principle:
Principle #26Copying

4Ease of operation

If the mannequin's face is replaced with the user's face picture, then user identification is improved, but the processing complexity increases

Engineering Contradiction:
Improveuser identification with mannequinVSAvoidface replacement processing
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The patent implements face replacement by copying the user's face image onto the mannequin's face region. This involves capturing or obtaining the user's face picture, processing it to match the mannequin's facial dimensions and orientation, and compositing it onto the three-dimensional model. While this does increase processing complexity compared to using a standard mannequin, the patent simplifies the process by using straightforward image mapping and blending techniques rather than complex 3D modeling or reconstruction methods.

Inventive Principle:
Principle #26Copying

Data Source

PatentUS12346959B2Display system and computer program product
Publication Date: 2025.07.01 SUETSUGU KATSUNORI
  • US12346959B2 patent drawing
  • US12346959B2 patent drawing
  • US12346959B2 patent drawing

AI summary

Provided is a display system and computer program product that are able to process a three-dimensional shape of a state in which a mannequin is wearing clothes, into an appearance desired by a user and to display the processed three-dimensional shape.A display system for selling clothes includes a wearing 3D generator that generates wearing 3D data representing a three-dimensional shape of a state in which a mannequin is wearing clothes by combining mannequin 3D data representing the mannequin in a three-dimensional shape using computer graphics and clothes 3D data representing the clothes in three-dimensional shapes using computer graphics and a display unit that displays the three-dimensional shape of the state in which the mannequin is wearing the clothes, on the basis of the wearing 3D data. The wearing 3D generator processes a portion specified by a user of the three-dimensional shape of the state in which the mannequin is wearing the clothes, into a specified form.