Weighted Mask Generation for Facial Expression Reproducibility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional weight mask generation devices have low reproducibility of facial expressions due to large errors in partial regions of interest, such as the mouth and eyebrows, when reconstructing a face shape from a face image.
Innovation Solution
A weight mask generation device that includes a displacement amount derivation unit, separation unit, normalization unit, and weight value generation unit to derive and normalize displacement amounts for each feature point across different deformation patterns of the face shape, improving reproducibility by generating weight values based on these normalized amounts and deformation degrees.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional weight masks assign uniform weights to all feature points regardless of facial expression changes, then the device complexity is low and ease of manufacture is high, but the manufacturing precision and reliability of facial expression reconstruction deteriorate
Solution Approach 1:
The patent segments the face shape into multiple partial regions (mouth, eyebrow, eye, etc.) and assigns different weight values to feature points based on their所属 partial region and deformation pattern. This segmentation allows the system to focus on regions with larger movement ranges that are more critical for facial expression reproduction, thereby improving manufacturing precision without requiring uniform high weights across all feature points.
Solution Approach 2:
The patent introduces dynamic weight adjustment based on deformation degrees. The weight mask generation unit calculates deformation degrees for each partial region and adjusts weight values dynamically according to the magnitude of deformation. This dynamic approach allows the system to adapt weights to actual facial expression changes, improving reproducibility while avoiding the need for static high weights everywhere.
2Reliability
If the weight mask focuses on partial regions with large movement ranges, then the reproducibility of facial expressions improves, but the measurement precision and difficulty of detecting and measuring other regions worsen
Solution Approach 1:
The patent applies local quality by assigning different weight values to different partial regions based on their specific characteristics. Regions with large movement ranges (mouth, eyebrow) receive higher weights, while regions with small movement ranges (eye, other areas) receive lower weights. This local differentiation allows the system to optimize for regions that most impact facial expression perception while maintaining adequate coverage of other regions.
Solution Approach 2:
The patent changes the weight parameter dynamically based on deformation degree calculations. By calculating the actual deformation magnitude for each partial region and adjusting weights accordingly, the system adapts to the specific facial expression being analyzed. This parameter change approach ensures that regions contributing most to expression variation receive appropriate attention while reducing the relative importance of regions with minimal movement.
Data Source
AI summary
A unit derives displacement amounts of feature points in an entire region of a face shape. A unit derives separated displacement amounts, each of which is a displacement amount of each feature point for every deformation pattern of a partial region of the face shape, based on the displacement amounts of the feature points in the entire region of the face shape and deformation degrees for every deformation pattern of the partial regions. A unit derives normalized and separated displacement amounts, which are obtained by normalizing the separated displacement amounts, for every deformation pattern of the partial regions. A unit generates weight values of the feature points in the entire region of the face shape based on the normalized and separated displacement amounts for every deformation pattern of the partial regions of the face shape and the deformation degrees for every deformation pattern of the partial regions.