3D Character Motion Retargeting via Synthetic Data Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The current methods for generating 3D character animation are time-consuming and cumbersome, requiring either manual definition of motion curves or complex motion capture processes, which are costly and inefficient, especially when trying to achieve complex motion data.
Innovation Solution
A system and method for generating synthetic motion data using pre-defined character motion models that combine and retarget motion data based on high-level descriptions, allowing for the creation of new motion sequences by leveraging analogies between pre-defined motions, and distributing processing across a shared server system and local computers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If manual definition of motion curves is used to animate 3D characters, then animation can be created without complex equipment, but the process is extremely time-consuming and laborious
Solution Approach 1:
The system captures real human motion data using motion capture equipment and creates digital copies of these motions that can be reused and retargeted to different 3D characters. This eliminates the need to manually create motion curves from scratch while reducing the need for repeated motion capture sessions.
Solution Approach 2:
Motion capture data is pre-recorded and stored in a library before animation needs to be created. This preliminary capture of diverse motions allows animators to quickly assemble and combine pre-captured motions rather than capturing or creating each motion sequence from scratch.
2Manufacturing precision
If motion capture is used to generate motion data, then realistic motion can be obtained, but complex equipment and actors are required increasing cost
Solution Approach 1:
The motion capture system is designed to capture diverse types of motions (walking, running, jumping, etc.) and store them in a universal library that can serve multiple animation projects and different 3D characters, eliminating the need for dedicated equipment setups for each specific motion type.
Solution Approach 2:
Instead of performing new motion capture for each animation need, the system creates and reuses digital copies of captured motions through retargeting technology, reducing the frequency and cost of actual motion capture sessions while maintaining high motion quality.
3Manufacturing precision
If motion capture is repeated to achieve desired motion, then accurate motion data can be obtained, but cost increases significantly
Solution Approach 1:
A comprehensive library of motion capture data is pre-recorded covering a wide range of motions and scenarios. This preliminary action ensures that when animation is needed, the required motions are already captured and can be selected and combined without needing to repeat expensive motion capture sessions.
Solution Approach 2:
The system combines multiple pre-captured motion sequences and blends them together to create complex animations. By merging existing motions through intelligent algorithms, the system achieves accurate and varied motion data without repeating motion capture for every animation scenario.
4Manufacturing precision
If manual editing of motion curves is attempted to correct captured motion, then desired motion can be obtained, but the process is difficult and time-consuming
Solution Approach 1:
The system uses retargeting technology to automatically copy and adapt motion capture data from one character to another, preserving the quality of the original capture while adapting it to different character proportions and anatomies without manual curve editing.
Solution Approach 2:
The retargeting system automatically adjusts motion parameters based on the target character's properties, transforming the motion data to fit different characters while maintaining the natural quality of the original capture, eliminating the need for manual parameter adjustment.
5Adaptability or versatility
If multiple animators need unique motion data, then each animator can create customized animations, but the workload and cost multiply
Solution Approach 1:
A single motion capture library serves multiple animators and projects simultaneously. The universal library contains diverse motions that can be selectively combined and retargeted to create unique animations for each animator without requiring separate motion capture sessions for each user.
Solution Approach 2:
The system enables multiple animators to work with the same motion library by allowing selective combination and blending of motions. Each animator can create unique animations by merging different pre-captured motions in their own way, maintaining productivity while achieving versatility.
Data Source
AI summary
Systems and methods are described for animating 3D characters using synthetic motion data generated by motion models in response to a high level description of a desired sequence of motion provided by an animator. In a number of embodiments, the synthetic motion data is streamed to a user device that includes a rendering engine and the user device renders an animation of a 3D character using the streamed synthetic motion data. In several embodiments, an animator can upload a custom model of a 3D character or a custom 3D character is generated by the server system in response to a high level description of a desired 3D character provided by the user and the synthetic motion data generated by the generative model is retargeted to animate the custom 3D character.