Saved in:
Bibliographic Details
Main Authors: Bello, Juan Luis Gonzalez, Yao, Xu, Whelan, Alex, Olszewski, Kyle, Kim, Hyeongwoo, Garrido, Pablo
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2504.07146
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910908073115648
author Bello, Juan Luis Gonzalez
Yao, Xu
Whelan, Alex
Olszewski, Kyle
Kim, Hyeongwoo
Garrido, Pablo
author_facet Bello, Juan Luis Gonzalez
Yao, Xu
Whelan, Alex
Olszewski, Kyle
Kim, Hyeongwoo
Garrido, Pablo
contents We present an implicit video representation for occlusions, appearance, and motion disentanglement from monocular videos, which we call Video SPatiotemporal Splines (VideoSPatS). Unlike previous methods that map time and coordinates to deformation and canonical colors, our VideoSPatS maps input coordinates into Spatial and Color Spline deformation fields $D_s$ and $D_c$, which disentangle motion and appearance in videos. With spline-based parametrization, our method naturally generates temporally consistent flow and guarantees long-term temporal consistency, which is crucial for convincing video editing. Using multiple prediction branches, our VideoSPatS model also performs layer separation between the latent video and the selected occluder. By disentangling occlusions, appearance, and motion, our method enables better spatiotemporal modeling and editing of diverse videos, including in-the-wild talking head videos with challenging occlusions, shadows, and specularities while maintaining an appropriate canonical space for editing. We also present general video modeling results on the DAVIS and CoDeF datasets, as well as our own talking head video dataset collected from open-source web videos. Extensive ablations show the combination of $D_s$ and $D_c$ under neural splines can overcome motion and appearance ambiguities, paving the way for more advanced video editing models.
format Preprint
id arxiv_https___arxiv_org_abs_2504_07146
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle VideoSPatS: Video SPatiotemporal Splines for Disentangled Occlusion, Appearance and Motion Modeling and Editing
Bello, Juan Luis Gonzalez
Yao, Xu
Whelan, Alex
Olszewski, Kyle
Kim, Hyeongwoo
Garrido, Pablo
Image and Video Processing
We present an implicit video representation for occlusions, appearance, and motion disentanglement from monocular videos, which we call Video SPatiotemporal Splines (VideoSPatS). Unlike previous methods that map time and coordinates to deformation and canonical colors, our VideoSPatS maps input coordinates into Spatial and Color Spline deformation fields $D_s$ and $D_c$, which disentangle motion and appearance in videos. With spline-based parametrization, our method naturally generates temporally consistent flow and guarantees long-term temporal consistency, which is crucial for convincing video editing. Using multiple prediction branches, our VideoSPatS model also performs layer separation between the latent video and the selected occluder. By disentangling occlusions, appearance, and motion, our method enables better spatiotemporal modeling and editing of diverse videos, including in-the-wild talking head videos with challenging occlusions, shadows, and specularities while maintaining an appropriate canonical space for editing. We also present general video modeling results on the DAVIS and CoDeF datasets, as well as our own talking head video dataset collected from open-source web videos. Extensive ablations show the combination of $D_s$ and $D_c$ under neural splines can overcome motion and appearance ambiguities, paving the way for more advanced video editing models.
title VideoSPatS: Video SPatiotemporal Splines for Disentangled Occlusion, Appearance and Motion Modeling and Editing
topic Image and Video Processing
url https://arxiv.org/abs/2504.07146