OmniMotionGPT: Animal Motion Generation with Limited Data

Mar 1, 2024·

Zhangsihao Yang

Mingyuan Zhou

Mengyi Shan

Bingbing Wen

Ziwei Xuan

Mitch Hill

Junjie Bai

Guo-Jun Qi

Yalin Wang

· 1 min read

Abstract

We propose OmniMotionGPT, a generative framework for animal motion synthesis that operates effectively in limited-data regimes, combining motion priors with powerful sequence modeling.

Type

Publication

CVPR 2024

We introduce OmniMotionGPT, a model for animal motion generation that leverages powerful sequence modeling to perform well even when training data is scarce.

Last updated on Mar 1, 2024

Motion Generation Computer Vision Generative Models

Authors

Bingbing Wen

← Laboratory-Scale AI: Open-Weight Models are Competitive with ChatGPT Even in Low-Resource Settings Jun 1, 2024

InfoVisDial: An Informative Visual Dialogue Dataset by Bridging Large Multimodal and Language Models Dec 1, 2023 →