Flexible Motion In-betweening with Diffusion Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cohan, Setareh, Tevet, Guy, Reda, Daniele, Peng, Xue Bin, van de Panne, Michiel
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913360944037888
author Cohan, Setareh
Tevet, Guy
Reda, Daniele
Peng, Xue Bin
van de Panne, Michiel
author_facet Cohan, Setareh
Tevet, Guy
Reda, Daniele
Peng, Xue Bin
van de Panne, Michiel
contents Motion in-betweening, a fundamental task in character animation, consists of generating motion sequences that plausibly interpolate user-provided keyframe constraints. It has long been recognized as a labor-intensive and challenging process. We investigate the potential of diffusion models in generating diverse human motions guided by keyframes. Unlike previous inbetweening methods, we propose a simple unified model capable of generating precise and diverse motions that conform to a flexible range of user-specified spatial constraints, as well as text conditioning. To this end, we propose Conditional Motion Diffusion In-betweening (CondMDI) which allows for arbitrary dense-or-sparse keyframe placement and partial keyframe constraints while generating high-quality motions that are diverse and coherent with the given keyframes. We evaluate the performance of CondMDI on the text-conditioned HumanML3D dataset and demonstrate the versatility and efficacy of diffusion models for keyframe in-betweening. We further explore the use of guidance and imputation-based approaches for inference-time keyframing and compare CondMDI against these methods.
format Preprint
id arxiv_https___arxiv_org_abs_2405_11126
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Flexible Motion In-betweening with Diffusion Models
Cohan, Setareh
Tevet, Guy
Reda, Daniele
Peng, Xue Bin
van de Panne, Michiel
Computer Vision and Pattern Recognition
Graphics
Machine Learning
Motion in-betweening, a fundamental task in character animation, consists of generating motion sequences that plausibly interpolate user-provided keyframe constraints. It has long been recognized as a labor-intensive and challenging process. We investigate the potential of diffusion models in generating diverse human motions guided by keyframes. Unlike previous inbetweening methods, we propose a simple unified model capable of generating precise and diverse motions that conform to a flexible range of user-specified spatial constraints, as well as text conditioning. To this end, we propose Conditional Motion Diffusion In-betweening (CondMDI) which allows for arbitrary dense-or-sparse keyframe placement and partial keyframe constraints while generating high-quality motions that are diverse and coherent with the given keyframes. We evaluate the performance of CondMDI on the text-conditioned HumanML3D dataset and demonstrate the versatility and efficacy of diffusion models for keyframe in-betweening. We further explore the use of guidance and imputation-based approaches for inference-time keyframing and compare CondMDI against these methods.
title Flexible Motion In-betweening with Diffusion Models
topic Computer Vision and Pattern Recognition
Graphics
Machine Learning
url https://arxiv.org/abs/2405.11126