Face Beautification via Displacement Vector Field
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing face beautification methods require manual adjustment, leading to low efficiency and poor beautification effects due to imprecise control over scale changes.
Innovation Solution
A method and apparatus that generate adjustment parameters based on a comparison between a target face and a reference face, using displacement vectors to adjust face shape and facial parts, improving control and precision in face beautification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual pushing and pulling adjustment is used, then the user can adjust face shape and facial parts, but the adjustment precision is low and efficiency is poor
Solution Approach 1:
The patent replaces the manual mechanical pushing and pulling operation with an automated computational system. The system uses facial feature point detection and displacement vector field technology to automatically calculate and apply adjustments, substituting manual mechanical interaction with algorithmic processing. This eliminates the imprecision of manual control while maintaining user-initiated beautification workflows.
Solution Approach 2:
The patent introduces displacement vector fields as an intermediary between user selection and face adjustment. Instead of direct manual manipulation, the system generates displacement vectors that represent the desired transformation, which then guide the automated adjustment process. This intermediary layer enables precise control over the adjustment magnitude and direction without requiring manual pixel-level manipulation.
2Measurement precision
If manual pushing and pulling is repeated multiple times, then adjustment precision may improve, but time consumption increases
Solution Approach 1:
The patent performs preliminary detection of facial feature points and pre-calculates displacement vectors based on the selected reference face and target face comparison. By preparing the adjustment parameters in advance through automated analysis rather than iterative manual adjustment, the system achieves precise results in a single operation, eliminating the time loss associated with repeated manual trials.
Solution Approach 2:
The system implements feedback by comparing the target face with the reference face to automatically determine the appropriate adjustment parameters. The displacement vector field provides continuous feedback information about the required transformation, enabling the system to self-correct and achieve precise adjustment without requiring multiple manual iteration cycles.
Data Source
AI summary
A face beautification method is performed by a computer device, the method including: obtaining a target face contained in a target image and a user-selected reference face; generating adjustment parameters corresponding to the target face based on a comparison of the target face and the reference face, the adjustment parameters comprising face shape adjustment parameters and facial part adjustment parameters; generating a displacement vector according to the face shape adjustment parameters and the facial part adjustment parameters, the displacement vector being used for representing a size change, a position change and an angle change in a face shape and facial parts of the target face during an adjustment process; adjusting the face shape and the facial parts of the target face according to the displacement vector; and displaying the adjusted target face.