Skinned Motion Retargeting with Dense Geometric Interaction Perception
Zijie Ye, Jia-Wei Liu, Jia Jia, Shikun Sun, Mike Zheng Shou

TL;DR
This paper introduces MeshRet, a novel motion retargeting framework that models dense geometric interactions to improve contact preservation and prevent interpenetration in skinned character animations.
Contribution
MeshRet directly models dense geometric interactions using a new DMI field, effectively addressing conflicts between skeleton motion and geometry correction in retargeting.
Findings
Achieves state-of-the-art performance on Mixamo and ScanRet datasets.
Effectively preserves contact and prevents interpenetration.
Outperforms existing methods in motion retargeting quality.
Abstract
Capturing and maintaining geometric interactions among different body parts is crucial for successful motion retargeting in skinned characters. Existing approaches often overlook body geometries or add a geometry correction stage after skeletal motion retargeting. This results in conflicts between skeleton interaction and geometry correction, leading to issues such as jittery, interpenetration, and contact mismatches. To address these challenges, we introduce a new retargeting framework, MeshRet, which directly models the dense geometric interactions in motion retargeting. Initially, we establish dense mesh correspondences between characters using semantically consistent sensors (SCS), effective across diverse mesh topologies. Subsequently, we develop a novel spatio-temporal representation called the dense mesh interaction (DMI) field. This field, a collection of interacting SCS feature…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
Taxonomy
TopicsAdvanced Vision and Imaging · Human Motion and Animation · Human Pose and Action Recognition
