Contact Preserving Shape Transfer For Rigging-Free Motion Retargeting

Abstract

Retargeting a motion from a source to a target character is an important problem in computer animation, as it allows to reuse existing rigged databases or transfer motion capture to virtual characters. Surface based pose transfer is a promising approach to avoid the trial-and-error process when controlling the joint angles. The main contribution of this paper is to investigate whether shape transfer instead of pose transfer would better preserve the original contextual meaning of the source pose. To this end, we propose an optimization-based method to deform the source shape+pose using three main energy functions: similarity to the target shape, body part volume preservation, and collision management (preserve existing contacts and prevent penetrations). The results show that our method is able to retarget complex poses, including several contacts, to very different morphologies. In particular, we introduce new contacts that are linked to the change in morphology, and which would be difficult to obtain with previous works based on pose transfer that aim at distance preservation between body parts. These preliminary results are encouraging and open several perspectives, such as decreasing computation time, and better understanding how to model pose and shape constraints. CCS CONCEPTS • Computing methodologies → Animation.

Reconstructing Human Body Mesh from Point Clouds by Adversarial GP Network

“…In practice, to accelerate the GP computation and improve the reconstruction precision, we apply the above statistical reasoning individually over body parts instead of the full body. We follow for that purpose [40] and segment the body topology into 19 small patches, including two patches on the elbow (see Fig. 1).…”

A functional skeleton transfer

“…Other approaches [Liu et al 2018], [Basset et al 2019] propose to transfer each pose of a 3D shape to another using a non-rigid registration. In these methods, a further adjustment is required to have a smooth animation from the sequence of poses obtained.…”

“…A similar work [Jin et al 2018] introduces the idea of aura mesh, a volume that surrounds the character and permits avoidance of self-collisions. The rigless methods [Basset et al 2019] do not need the post-processing angle correction, but they require additional steps to obtain an animation-ready shape as output. In [Avril et al 2016] the authors compute a pointwise matching between the two shapes to transfer all the properties of an animation setup.…”

Contact-Aware Retargeting of Skinned Motion

“…Ho and Shum [15] adapt robot motions with a spacetime optimization strategy that prevents self-collisions. More recently, Basset et al [1,7] proposed a constrained optimization algorithm with attraction and repulsion terms to preserve contacts and avoid interpenetration. These works, however, require exact vertex correspondences between the input and target meshes, expensive constrained optimization over the entire sequence, and smoothing as a post-process.…”