Surface based motion retargeting by preserving spatial relationship

Abstract

Retargeting motion from one character to another is a key process in computer animation. It enables to reuse animations designed for a character to animate another one, or to make performance-driven be faithful to what has been performed by the user. Previous work mainly focused on retargeting skeleton animations whereas the contextual meaning of the motion is mainly linked to the relationship between body surfaces, such as the contact of the palm with the belly. In this paper we propose a new context-aware motion retargeting framework, based on deforming a target character to mimic a source character poses using harmonic mapping. We also introduce the idea of Context Graph: modeling local interactions between surfaces of the source character, to be preserved in the target character, in order to ensure fidelity of the pose. In this approach, no rigging is required as we directly manipulate the surfaces, which makes the process totally automatic. Our results demonstrate the relevance of this automatic rigging-less approach on motions with complex contacts and interactions between the character's surface.

Contact Preserving Shape Transfer For Rigging-Free Motion Retargeting

“…The main contribution of the paper is to explore transferring the shape instead of the pose, assuming that it would better preserve the contextual meaning of the source shape+pose. Whereas other methods [Jin et al 2018;Liu et al 2018] preserve the distances between surfaces, we allow surface interactions (e.g. distances) to change during the transfer because the shape is changing.…”

“…Recent works have added constraints to encode the pose as spatial relationships between the surfaces of the body. Liu et al [Liu et al 2018] introduced the idea of context graph, an extension of interaction mesh [Ho et al 2010] to deal with body surface instead of joint centers. In context graph, nodes are points placed on the body surface and edges model distance constraints in-between these points.…”

“…Thus, changing the morphology of a character while keeping the same pose might result in impossible and/or semantically different poses. Previous works [Ho et al 2010;Jin et al 2018;Liu et al 2018] have solved this problem by deforming the pose to preserve the distances between interacting surfaces or joints. We argue that distances between surfaces are dependent of the character's morphology.…”

“…For our experiments, to have point to point correspondence, all our meshes follow the SMPL [Loper et al 2015] template (6890 vertices and 13776 faces), segmented in 17 body parts (see Figure 3a). We use meshes from SMPL, Faust [Bogo et al 2014], Dyna , and from Liu et al [Liu et al 2018]. All experiments were run on a PC with an Intel Xeon CPU and 16GB of RAM.…”

“…We argue that our method proposes an alternative to such distance preserving methods [Ho et al 2010;Jin et al 2018;Liu et al 2018], and is a first step in taking into account the correlation between shape and pose in motion retargeting.…”