Enregistré dans:
Détails bibliographiques
Auteurs principaux: Zhao, Jinjing, Wei, Fangyun, Liu, Zhening, Zhang, Hongyang, Xu, Chang, Lu, Yan
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2512.15716
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866912771360161792
author Zhao, Jinjing
Wei, Fangyun
Liu, Zhening
Zhang, Hongyang
Xu, Chang
Lu, Yan
author_facet Zhao, Jinjing
Wei, Fangyun
Liu, Zhening
Zhang, Hongyang
Xu, Chang
Lu, Yan
contents Existing video generation models struggle to maintain long-term spatial and temporal consistency due to the dense, high-dimensional nature of video signals. To overcome this limitation, we propose Spatia, a spatial memory-aware video generation framework that explicitly preserves a 3D scene point cloud as persistent spatial memory. Spatia iteratively generates video clips conditioned on this spatial memory and continuously updates it through visual SLAM. This dynamic-static disentanglement design enhances spatial consistency throughout the generation process while preserving the model's ability to produce realistic dynamic entities. Furthermore, Spatia enables applications such as explicit camera control and 3D-aware interactive editing, providing a geometrically grounded framework for scalable, memory-driven video generation.
format Preprint
id arxiv_https___arxiv_org_abs_2512_15716
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Spatia: Video Generation with Updatable Spatial Memory
Zhao, Jinjing
Wei, Fangyun
Liu, Zhening
Zhang, Hongyang
Xu, Chang
Lu, Yan
Computer Vision and Pattern Recognition
Artificial Intelligence
Existing video generation models struggle to maintain long-term spatial and temporal consistency due to the dense, high-dimensional nature of video signals. To overcome this limitation, we propose Spatia, a spatial memory-aware video generation framework that explicitly preserves a 3D scene point cloud as persistent spatial memory. Spatia iteratively generates video clips conditioned on this spatial memory and continuously updates it through visual SLAM. This dynamic-static disentanglement design enhances spatial consistency throughout the generation process while preserving the model's ability to produce realistic dynamic entities. Furthermore, Spatia enables applications such as explicit camera control and 3D-aware interactive editing, providing a geometrically grounded framework for scalable, memory-driven video generation.
title Spatia: Video Generation with Updatable Spatial Memory
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2512.15716