Digital Human Action Redirection Using Body-Segment Collision Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing action redirection methods for digital humans in the metaverse are time-consuming and lack real-time performance, leading to unrealistic simulations due to clipping issues when actions from a source character are applied to a target character with body or clothing differences.
Innovation Solution
Segment the target character into body segments, model each segment using a coordinate system, perform collision detection, and adjust distances and action sequences to ensure geometric accuracy and prevent clipping, while transferring action semantics from the source character to the target character.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual correction is used to redirect clipping actions, then action quality is improved, but time consumption increases and real-time performance deteriorates
Solution Approach 1:
The patent replaces the manual mechanical correction process with an automated computational system. A computing device automatically performs collision detection on geometric models of body segments and redirects clipping actions through algorithmic processing, eliminating the need for manual frame-by-frame correction while maintaining action quality.
Solution Approach 2:
The system enables self-service by allowing the action redirection process to automatically detect and correct clipping issues without human intervention. The computing device autonomously processes the action sequence, identifies collisions between body segments, and generates corrected actions, making the system self-sufficient for quality improvement.
2Measurement precision
If geometric modeling and collision detection are performed for each body segment, then clipping detection accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent divides the target character into multiple geometric models corresponding to different body segments (head, torso, limbs, etc.). Each body segment is independently modeled and subjected to collision detection, enabling precise clipping detection at the segment level while maintaining overall system manageability through modular processing.
Data Source
AI summary
An action redirection method includes: segmenting a target character based on semantics of different parts of the target character, to obtain a plurality of first body segments related to the target character; modeling each first body segment based on a first coordinate system marked on the first body segment, to obtain a plurality of first geometric models, where one first geometric model represents a shape of one body segment; performing collision detection on the plurality of first geometric models; and when first geometric models collide with each other, adjusting a distance between the first geometric models that collide with each other, and adjusting an action sequence of the target character based on an adjusted distance. Thus, a coordinate system is constructed for a body segment of the target character, and a geometric model is output through modeling, so that accurate geometric expressing of the body segment is implemented.