Unifying Video Tasks via Spatiotemporal Analogy

arXiv:2609.33935 · cs.CV, cs.AI · Submitted 2026-09-27 · Read on arXiv

cs.CV, cs.AI

Submitted: 2026-09-27

Updated: 2026-09-30

Comments: Project page: https://iandrover.github.io/video_analogy/

Project page: https://iandrover.github.io/video_analogy

License: http://creativecommons.org/licenses/by/4.0/

The gist: Adapting video models to new tasks typically requires dedicated data curation and fine-tuning.

Terminology

Abstract

Adapting video models to new tasks typically requires dedicated data curation and fine-tuning. While visual analogy provides a training-free alternative by specifying tasks in-context, it remains restricted to the image domain. To explore whether analogy-based methods can unify diverse video tasks and generalize to out-of-distribution scenarios, we introduce ViGeo, a framework that extends visual in-context learning to the video domain via spatiotemporal canvas completion. Evaluated on a diverse task taxonomy with a strict train-test split, ViGeo generalizes to unseen video manipulations and zero-shot modalities (e.g., event cameras). Finally, we identify task internalization, where a query format associated with a pretrained task overrides the demonstration, and show that this shortcut can be removed with a small amount of task-unrelated data, highlighting the need to decorrelate prompt format from task identity.

Sources

Related papers