Guardado en:
Detalles Bibliográficos
Autores principales: Hua, Yu, Liu, Weiming, Xu, Gui, Hou, Yaqing, Ong, Yew-Soon, Zhang, Qiang
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2505.00998
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912357625626624
author Hua, Yu
Liu, Weiming
Xu, Gui
Hou, Yaqing
Ong, Yew-Soon
Zhang, Qiang
author_facet Hua, Yu
Liu, Weiming
Xu, Gui
Hou, Yaqing
Ong, Yew-Soon
Zhang, Qiang
contents Human motion synthesis aims to generate plausible human motion sequences, which has raised widespread attention in computer animation. Recent score-based generative models (SGMs) have demonstrated impressive results on this task. However, their training process involves complex curvature trajectories, leading to unstable training process. In this paper, we propose a Deterministic-to-Stochastic Diverse Latent Feature Mapping (DSDFM) method for human motion synthesis. DSDFM consists of two stages. The first human motion reconstruction stage aims to learn the latent space distribution of human motions. The second diverse motion generation stage aims to build connections between the Gaussian distribution and the latent space distribution of human motions, thereby enhancing the diversity and accuracy of the generated human motions. This stage is achieved by the designed deterministic feature mapping procedure with DerODE and stochastic diverse output generation procedure with DivSDE.DSDFM is easy to train compared to previous SGMs-based methods and can enhance diversity without introducing additional training parameters.Through qualitative and quantitative experiments, DSDFM achieves state-of-the-art results surpassing the latest methods, validating its superiority in human motion synthesis.
format Preprint
id arxiv_https___arxiv_org_abs_2505_00998
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Deterministic-to-Stochastic Diverse Latent Feature Mapping for Human Motion Synthesis
Hua, Yu
Liu, Weiming
Xu, Gui
Hou, Yaqing
Ong, Yew-Soon
Zhang, Qiang
Computer Vision and Pattern Recognition
Human motion synthesis aims to generate plausible human motion sequences, which has raised widespread attention in computer animation. Recent score-based generative models (SGMs) have demonstrated impressive results on this task. However, their training process involves complex curvature trajectories, leading to unstable training process. In this paper, we propose a Deterministic-to-Stochastic Diverse Latent Feature Mapping (DSDFM) method for human motion synthesis. DSDFM consists of two stages. The first human motion reconstruction stage aims to learn the latent space distribution of human motions. The second diverse motion generation stage aims to build connections between the Gaussian distribution and the latent space distribution of human motions, thereby enhancing the diversity and accuracy of the generated human motions. This stage is achieved by the designed deterministic feature mapping procedure with DerODE and stochastic diverse output generation procedure with DivSDE.DSDFM is easy to train compared to previous SGMs-based methods and can enhance diversity without introducing additional training parameters.Through qualitative and quantitative experiments, DSDFM achieves state-of-the-art results surpassing the latest methods, validating its superiority in human motion synthesis.
title Deterministic-to-Stochastic Diverse Latent Feature Mapping for Human Motion Synthesis
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2505.00998