arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2501.06035cs.CV
收藏

非各向同性高斯扩散用于真实三维人体运动预测

Nonisotropic Gaussian Diffusion for Realistic 3D Human Motion Prediction

Technical University of Munich(慕尼黑技术大学);Munich Center for Machine Learning(慕尼黑机器学习中心)

Cecilia Curreli, Dominik Muhle, Abhishek Saroha, Zhenzhang Ye, Riccardo Marin, Daniel Cremers

AI总结:

提出SkeletonDiffusion,一种非各向同性高斯扩散的潜在扩散模型,用于真实3D人体运动预测,避免肢体伪影并优于现有基线。

摘要:

概率性人体运动预测旨在从过去的观测中预测多种可能的未来运动。尽管当前方法报告了高多样性和真实性,它们常常生成带有未检测到的肢体拉伸和抖动伪影的运动。为解决此问题,我们引入了SkeletonDiffusion,一种在其架构和训练中嵌入对人体显式归纳偏置的潜在扩散模型。我们的模型采用一种新颖的非各向同性高斯扩散公式进行训练,该公式与人体骨骼的自然运动学结构对齐。结果表明,我们的方法优于传统的各向同性替代方案,一致地生成真实预测,同时避免肢体变形等伪影。此外,我们识别了常用多样性指标的一个局限性,该指标可能无意中偏向于在相同序列内产生不一致肢体长度的模型。SkeletonDiffusion在真实世界数据集上树立了新基准,在多种评估指标上优于各种基线。访问我们的项目页面:https://ceveloper.github.io/publications/skeletondiffusion/ 。

英文摘要:

Probabilistic human motion prediction aims to forecast multiple possible future movements from past observations. While current approaches report high diversity and realism, they often generate motions with undetected limb stretching and jitter. To address this, we introduce SkeletonDiffusion, a latent diffusion model that embeds an explicit inductive bias on the human body within its architecture and training. Our model is trained with a novel nonisotropic Gaussian diffusion formulation that aligns with the natural kinematic structure of the human skeleton. Results show that our approach outperforms conventional isotropic alternatives, consistently generating realistic predictions while avoiding artifacts such as limb distortion. Additionally, we identify a limitation in commonly used diversity metrics, which may inadvertently favor models that produce inconsistent limb lengths within the same sequence. SkeletonDiffusion sets a new benchmark on real-world datasets, outperforming various baselines across multiple evaluation metrics. Visit our project page at https://ceveloper.github.io/publications/skeletondiffusion/ .

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑