Ethnicity-Aware Facial Aging Simulation for Realistic Progression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing age prediction and progression systems fail to accurately account for ethnic-dependent changes in face color, shape, and texture, lacking realism in simulating age appearance.
Innovation Solution
Utilizing statistical ethnic models to alter facial features, including shape, color, and texture, based on ethnicity, age, and gender inputs to create realistic age simulations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional modeling techniques are used for age simulation, then the aging process can be performed, but the realism of gradual shape, color, and texture changes is lost
Solution Approach 1:
The patent applies parameter changes by using statistical models that define ethnic-dependent parameters for shape, color, and texture changes across different ages. These parameters are derived from training images of individuals from different ethnic groups, allowing the system to adjust facial features according to specific ethnic aging patterns while maintaining realism throughout the aging process
Solution Approach 2:
The patent segments the aging process into three distinct components: shape changes, color changes, and texture changes. Each component is modeled separately using dedicated statistical models (shape model, color model, and texture model), allowing for precise control and realistic simulation of each aspect of aging independently while maintaining overall coherence
2Adaptability or versatility
If generic age prediction systems are used, then age progression can be performed, but ethnic differences in aging are not accounted for
Solution Approach 1:
The patent implements local quality by creating ethnicity-specific statistical models that capture local characteristics of different ethnic groups. Each ethnic group has its own set of parameters derived from training images, allowing the system to apply appropriate aging patterns specific to each ethnicity rather than using a one-size-fits-all approach
Solution Approach 2:
The patent applies preliminary action by pre-training statistical models on large datasets of facial images from different ethnic groups before actual age simulation. The training process pre-computes mean images, covariance matrices, and eigenfaces for each ethnic group, enabling accurate and efficient ethnic-dependent aging without requiring complex computations during the actual simulation process
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
Methods and systems for age appearance simulation of a consumer are provided. At least one embodiment of a method includes receiving an image of the consumer (including a face of the consumer), determining an ethnicity of the consumer, determining an age of the patient, and determining a desired simulated age of the consumer. An altered image to represent the desired simulated age of the consumer may then be created, where altering the image includes utilizing a statistical ethnic aging model to alter at least one of a shape of the face, a color of the face, and a texture of the face. The altered image the first altered image may be provided for display to the consumer.