Salvato in:
Dettagli Bibliografici
Autori principali: Tripathi, Shashank, Taheri, Omid, Lassner, Christoph, Black, Michael J., Holden, Daniel, Stoll, Carsten
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2409.03944
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916672326074368
author Tripathi, Shashank
Taheri, Omid
Lassner, Christoph
Black, Michael J.
Holden, Daniel
Stoll, Carsten
author_facet Tripathi, Shashank
Taheri, Omid
Lassner, Christoph
Black, Michael J.
Holden, Daniel
Stoll, Carsten
contents Generating realistic human motion is essential for many computer vision and graphics applications. The wide variety of human body shapes and sizes greatly impacts how people move. However, most existing motion models ignore these differences, relying on a standardized, average body. This leads to uniform motion across different body types, where movements don't match their physical characteristics, limiting diversity. To solve this, we introduce a new approach to develop a generative motion model based on body shape. We show that it's possible to train this model using unpaired data by applying cycle consistency, intuitive physics, and stability constraints, which capture the relationship between identity and movement. The resulting model generates diverse, physically plausible, and dynamically stable human motions that are both quantitatively and qualitatively more realistic than current state-of-the-art methods. More details are available on our project page https://CarstenEpic.github.io/humos/.
format Preprint
id arxiv_https___arxiv_org_abs_2409_03944
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle HUMOS: Human Motion Model Conditioned on Body Shape
Tripathi, Shashank
Taheri, Omid
Lassner, Christoph
Black, Michael J.
Holden, Daniel
Stoll, Carsten
Computer Vision and Pattern Recognition
Artificial Intelligence
Generating realistic human motion is essential for many computer vision and graphics applications. The wide variety of human body shapes and sizes greatly impacts how people move. However, most existing motion models ignore these differences, relying on a standardized, average body. This leads to uniform motion across different body types, where movements don't match their physical characteristics, limiting diversity. To solve this, we introduce a new approach to develop a generative motion model based on body shape. We show that it's possible to train this model using unpaired data by applying cycle consistency, intuitive physics, and stability constraints, which capture the relationship between identity and movement. The resulting model generates diverse, physically plausible, and dynamically stable human motions that are both quantitatively and qualitatively more realistic than current state-of-the-art methods. More details are available on our project page https://CarstenEpic.github.io/humos/.
title HUMOS: Human Motion Model Conditioned on Body Shape
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2409.03944