Energy-Based Pose Unfolding and Interpolation for 3D Articulated Characters

IEEE Trans. Visual. Comput. Graphics

Komura

2015

Self Cite

Abstract-In this paper, we present a local motion planning algorithm for character animation. We focus on motion planning between two distant postures where linear interpolation leads to penetrations. Our framework has two stages. The motion planning problem is first solved as a Boundary Value Problem (BVP) on an energy graph which encodes penetrations, motion smoothness and user control. Having established a mapping from the configuration space to the energy graph, a fast and robust local motion planning algorithm is introduced to solve the BVP to generate motions that could only previously be computed by global planning methods. In the second stage, a projection of the solution motion onto a constraint manifold is proposed for more user control. Our method can be integrated into current keyframing techniques. It also has potential applications in motion planning problems in robotics.

“…Relation to Previous Work Wang and Komura [23] also use a repulsive energy function to do motion planning between distant postures. Our energy function is an augmented version of theirs.…”

Section: Discussion and Limitationsmentioning

confidence: 99%

“…The most similar work to ours is [23], which uses linkage unfolding to plan the motions. However, their method is solely based on linkage folding/unfolding.…”

Section: Path-planning Of Close Interactionsmentioning

confidence: 99%

See 1 more Smart Citation

An Energy-Driven Motion Planning Method for Two Distant Postures

IEEE Trans. Visual. Comput. Graphics

Komura

2015

Self Cite

“…In-between motion synthesis can be regarded as motion planning [Arikan and Forsyth 2002;Beaudoin et al 2008;Levine et al 2012;Safonova and Hodgins 2007;Wang and Komura 2011;Wang et al 2013], which solves complex optimization problems with various constraints. Data-driven methods are faster and produce more natural motions by searching in structured data, such as motion graphs [Kovar et al 2008;Min and Chai 2012;Shen et al 2017], but there is a trade-off between the diversity of animation and the space complexity.…”

Section: Related Work 21 In-between Motion Generationmentioning

confidence: 99%

RSMT: Real-time Stylized Motion Transition for Characters

Tang

Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Proceedings

et al. 2023

Figure 1: Our method generates an in-between motion sequence with a high leg lifting style (blue) between target frames (orange). Given a target frame, a desired transition duration and the required style, the character can dynamically adjust motions to reach the target with the desired style. Specifically, the character in the white box needs bigger steps to reach the target within the specified duration, while the character in the orange box needs smaller steps to reduce the speed for the same reason.

“…Existing methods formulate in-between motions in various fashions. In the early days, in-between motions were often formulated as a motion planning problem [Arikan and Forsyth 2002;Beaudoin et al 2008;Levine et al 2012;Safonova and Hodgins 2007;Wang and Komura 2011], where fairly sophisticated motions can be synthesized. Complex optimization problems [Chai and Hodgins 2007] are formed with respect to various constraints such as contact and control input, leading to slow computation which is impractical for the animators and impossible for real-time applications.…”

Section: Related Workmentioning

confidence: 99%

Real-time controllable motion transition for characters

Tang

et al. 2022

ACM Trans. Graph.

Self Cite

Real-time in-between motion generation is universally required in games and highly desirable in existing animation pipelines. Its core challenge lies in the need to satisfy three critical conditions simultaneously: quality, controllability and speed , which renders any methods that need offline computation (or post-processing) or cannot incorporate (often unpredictable) user control undesirable. To this end, we propose a new real-time transition method to address the aforementioned challenges. Our approach consists of two key components: motion manifold and conditional transitioning. The former learns the important low-level motion features and their dynamics; while the latter synthesizes transitions conditioned on a target frame and the desired transition duration. We first learn a motion manifold that explicitly models the intrinsic transition stochasticity in human motions via a multi-modal mapping mechanism. Then, during generation, we design a transition model which is essentially a sampling strategy to sample from the learned manifold, based on the target frame and the aimed transition duration. We validate our method on different datasets in tasks where no post-processing or offline computation is allowed. Through exhaustive evaluation and comparison, we show that our method is able to generate high-quality motions measured under multiple metrics. Our method is also robust under various target frames (with extreme cases).