IBMM++: Improved and versatile motion generation with implicit Bézier curves
In this work, we introduce extended version (IBMM++), we further improve constraint accuracy, reduce foot sliding, and enable motion extension, orientation constraints, longer sequences, and stable iterative editing through architectural updates, a revised tangent formulation, a randomized constraint sampling strategy, and dedicated inference-time methods, validated on a larger internal dataset with more varied motions.
October 16, 2026Computers & Graphics (10/20/2026 issue)
Authors
Dhruv Agrawal (DisneyResearch|Studios/ETH Zurich)
Luca Vögeli (DisneyResearch|Studios)
Dominik Borer (DisneyResearch|Studios)
Robert W. Sumner (DisneyResearch|Studios/ETH Zurich)
Martin Guay (DisneyResearch|Studios)
Jakob Buhmann (DisneyResearch|Studios)
IBMM++: Improved and versatile motion generation with implicit Bézier curves
Abstract
Creating high-quality character animation remains an intricate and cumbersome process. Recently, diffusion models have unlocked the ability to generate diverse movements from high-level condition signals such as text. For artist-friendly control, motion diffusion leveraging Bézier curves have been shown to allow precise joint-level conditioning. Yet, these works have been limited to joints at a fixed temporal stride, while animators require more temporal flexibility when keyframing or manipulating tangents to achieve animation principles such as easing in & out. In Vögeli et al. we presented a new Implicit Bézier Motion Model (IBMM), which during training is exposed to all possible configurations of control points, enabling control at arbitrary timings. In addition, we introduced a new quantitative measure of ease-in and-out, which leads to a novel condition over the motion generation process to reflect this artistic principle. In this extended version (IBMM++), we further improve constraint accuracy, reduce foot sliding, and enable motion extension, orientation constraints, longer sequences, and stable iterative editing through architectural updates, a revised tangent formulation, a randomized constraint sampling strategy, and dedicated inference-time methods, validated on a larger internal dataset with more varied motions.
Copyright Notice
The documents contained in these directories are included by the contributing authors as a means to ensure timely dissemination of scholarly and technical work on a non-commercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author’s copyright. These works may not be reposted without the explicit permission of the copyright holder.
点赞
收藏
点踩
分享