Avatar Facial Expression Mapping via Component Event Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods for applying facial expressions to avatars often result in unnatural expressions when the user's face is dissimilar to the avatar's face, leading to unsuitable facial mappings.
Innovation Solution
An image processing method that extracts landmark information from an input image, maps the object to a virtual object, determines component values, selects a corresponding event from predefined events based on these values, and applies the event to the virtual object, ensuring a natural and suitable facial expression is applied.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If feature point locations tracked from a user's face are applied directly to an avatar, then the user's facial expression is accurately transferred, but the avatar exhibits unnatural expressions when the faces are dissimilar
Solution Approach 1:
The patent introduces an intermediate processing layer that includes component value determination and event selection modules. Instead of directly mapping feature points, the system determines values of facial components (eyes, nose, mouth, eyebrows, face angle), selects appropriate events based on combinations of these values, and applies them to the virtual object. This intermediary process adapts the user's facial expression to suit the avatar's characteristics, resolving the contradiction between accurate transfer and natural appearance.
Solution Approach 2:
The system changes the parameters used for facial expression mapping from direct feature point coordinates to component-based descriptors (eye shape, nose shape, mouth shape, eyebrow position, face angle). By transforming the representation parameters and selecting events based on combinations of these parameter values, the system achieves both accurate expression transfer and natural avatar appearance.
2Reliability
If multiple component values are determined and events are selected based on their combinations, then the facial expression becomes more natural for the virtual object, but the processing complexity increases
Solution Approach 1:
The patent segments the facial expression processing into distinct components: determining values for individual facial elements (eyes, nose, mouth, eyebrows, face angle), selecting events based on combinations of these component values, and applying the selected events. This segmentation allows the complex task of natural expression mapping to be broken down into manageable steps, reducing processing complexity while maintaining expression naturalness.
Data Source
AI summary
Provided is an image processing method and system. More particularly, provided are an image processing method and system that may map a facial expression of a user, a specific event, etc., to a character in real time by mapping an object such as a face of a user included in an input image to a virtual object such as a face of a specific character, and by applying an event corresponding to a combination of components defined for the object to the virtual object, and a non-transitory computer-readable recording medium that, when executed by a processor, cause the processor to implement the image processing method in conjunction with a computer device.