← ArchivePaper2015
Motion Retargeting for Crowd Simulation
Presents real-time motion retargeting and interactive motion blending within the Golaem Crowd plugin, reducing asset count for production crowd shots.
Abstract
Motion retargeting for crowd simulation using an intermediate skeleton representation (Golaem Skeleton) based on hierarchical bone chains. The system enables real-time retargeting of animations across different character morphologies by defining skeleton-motion mappings and using inverse kinematics algorithms (analytic solution and FABRIK) with caching optimization to support thousands of characters in production crowds.
How to read this
- Category
- Production system: crowd motion retargeting (Golaem)
- Contributions
- Introduces an intermediate skeleton representation (Golaem Skeleton) of hierarchical bone chains for mapping motion across differing character morphologies.
- Performs real-time retargeting using analytic and FABRIK inverse-kinematics solvers with caching to scale to thousands of characters.
- Reduces asset count for production crowd shots via interactive motion blending inside the Golaem Crowd plugin.
- Context
- An applied, crowd-scale extension of Gleicher's Retargeting Motion to New Characters, adapting per-character retargeting to thousands of agents in production.Builds on: Retargeting Motion to New Characters
- Correctness
- An engineering/production paper: validated by use in crowd shots rather than formal benchmarks, so the intermediate-skeleton mapping and IK choices trade some per-character fidelity for the throughput needed at crowd scale.
- Clarity
- Accessible and practical; a single pass conveys the pipeline, with a second pass only needed for the skeleton-mapping and caching details.
- How to read it
- First pass for the intermediate-skeleton and caching strategy; revisit the IK-solver specifics only if implementing a crowd retargeting pipeline.
Builds on
Built upon by
Nothing yet.
Related work
- Mobilizing Mocap, Motion Blending, and Mayhem: Rig Interoperability for Crowd Simulation on Incredibles 2 2018 / SIGGRAPH
- Normalized Euclidean Distance Matrices for Human Motion Retargeting 2017 / MIG
- Robust Marker Trajectory Repair for MOCAP Using Kinematic Reference 2019 / MIG
- Dog Code: Human to Quadruped Embodiment Using Shared Codebooks 2024 / MIG
Keywords
This page summarises the entry and links to its original source. The archive never hosts or redistributes the publication itself.Show it in the full archive list →