WorldPack: Dynamic Frame Compression for Long-context Video World Modeling

Abstract:Video world models have attracted significant attention for their ability to produce high-fidelity future visual observations conditioned on past observations and navigation actions. However, achieving temporally and spatially consistent generation over long horizons remains an open challenge: existing approaches either compress past frames using generic importance schedules that do not exploit 3D viewpoint geometry, or retrieve only a handful of spatially relevant frames without increasing the total amount of retained history. In this paper, we propose WorldPack, a video world model that introduces spatially-aware compressed memory to address both limitations simultaneously. The key insight is that compression rates should not be uniform or temporally determined, but should instead be dynamically allocated based on 3D spatial relevance to the current this http URL achieves this through two tightly coupled mechanisms: trajectory packing, which fits substantially more historical frames into a fixed-length context through hierarchical frame compression, and geometric selection, which leverages camera pose information and field-of-view overlap to assign lower compression to spatially important frames and higher compression to less relevant ones. Together, these mechanisms expand the effective context from 4 to 22 frames with moderate computational overhead: trajectory packing increases diffusion-model inference time by 16%, while FoV-based geometric selection introduces an additional candidate-dependent cost. We evaluate WorldPack on LoopNav, a Minecraft benchmark for long-horizon spatial consistency, and conduct comprehensive experiments on the RECON, real-world navigation dataset, across multiple evaluation protocols. WorldPack outperforms strong baselines, Oasis, Mineworld, DIAMOND, NWM, with pronounced gains on spatial reasoning tasks requiring recall of distant observations.

Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)

arXiv:2512.02473 [cs.CV]

(or arXiv:2512.02473v2 [cs.CV] for this version)

https://doi.org/10.48550/arXiv.2512.02473 Focus to learn more arXiv-issued DOI via DataCite

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.