English

GeCo: Evaluating Geometric Consistency for Video Generation via Motion and Structure

Computer Vision and Pattern Recognition 2026-04-24 v2

Abstract

We introduce GeCo, a geometry-grounded metric for jointly detecting geometric deformation and occlusion-inconsistency artifacts in static scenes. By fusing residual motion and depth priors, GeCo produces interpretable, dense consistency maps that reveal these artifacts. We use GeCo to systematically benchmark recent video generation models, uncovering common failure modes, and further employ it as a training-free guidance loss to reduce deformation artifacts during video generation.

Cite

@article{arxiv.2512.22274,
  title  = {GeCo: Evaluating Geometric Consistency for Video Generation via Motion and Structure},
  author = {Leslie Gu and Junhwa Hur and Charles Herrmann and Fangneng Zhan and Todd Zickler and Deqing Sun and Hanspeter Pfister},
  journal= {arXiv preprint arXiv:2512.22274},
  year   = {2026}
}
R2 v1 2026-07-01T08:42:01.675Z