Saved in:
Bibliographic Details
Main Authors: Shi, Xu, Yao, Wei, Luo, Chuanchen, Peng, Junran, Zhang, Hongwen, Sun, Yunlian
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2312.02772
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916507875803136
author Shi, Xu
Yao, Wei
Luo, Chuanchen
Peng, Junran
Zhang, Hongwen
Sun, Yunlian
author_facet Shi, Xu
Yao, Wei
Luo, Chuanchen
Peng, Junran
Zhang, Hongwen
Sun, Yunlian
contents Recently, significant progress has been made in text-based motion generation, enabling the generation of diverse and high-quality human motions that conform to textual descriptions. However, generating motions beyond the distribution of original datasets remains challenging, i.e., zero-shot generation. By adopting a divide-and-conquer strategy, we propose a new framework named Fine-Grained Human Motion Diffusion Model (FG-MDM) for zero-shot human motion generation. Specifically, we first parse previous vague textual annotations into fine-grained descriptions of different body parts by leveraging a large language model. We then use these fine-grained descriptions to guide a transformer-based diffusion model, which further adopts a design of part tokens. FG-MDM can generate human motions beyond the scope of original datasets owing to descriptions that are closer to motion essence. Our experimental results demonstrate the superiority of FG-MDM over previous methods in zero-shot settings. We will release our fine-grained textual annotations for HumanML3D and KIT.
format Preprint
id arxiv_https___arxiv_org_abs_2312_02772
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle FG-MDM: Towards Zero-Shot Human Motion Generation via ChatGPT-Refined Descriptions
Shi, Xu
Yao, Wei
Luo, Chuanchen
Peng, Junran
Zhang, Hongwen
Sun, Yunlian
Computer Vision and Pattern Recognition
Recently, significant progress has been made in text-based motion generation, enabling the generation of diverse and high-quality human motions that conform to textual descriptions. However, generating motions beyond the distribution of original datasets remains challenging, i.e., zero-shot generation. By adopting a divide-and-conquer strategy, we propose a new framework named Fine-Grained Human Motion Diffusion Model (FG-MDM) for zero-shot human motion generation. Specifically, we first parse previous vague textual annotations into fine-grained descriptions of different body parts by leveraging a large language model. We then use these fine-grained descriptions to guide a transformer-based diffusion model, which further adopts a design of part tokens. FG-MDM can generate human motions beyond the scope of original datasets owing to descriptions that are closer to motion essence. Our experimental results demonstrate the superiority of FG-MDM over previous methods in zero-shot settings. We will release our fine-grained textual annotations for HumanML3D and KIT.
title FG-MDM: Towards Zero-Shot Human Motion Generation via ChatGPT-Refined Descriptions
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2312.02772