English

SCP: Spatial Causal Prediction in Video

Computer Vision and Pattern Recognition 2026-04-07 v2

Abstract

Spatial reasoning, the ability to understand spatial relations, causality, and dynamic evolution, is central to human intelligence and essential for real-world applications such as autonomous driving and robotics. Existing studies, however, primarily assess models on visible spatio-temporal understanding, overlooking their ability to infer unseen past or future spatial states. In this work, we introduce Spatial Causal Prediction (SCP), a new task paradigm that challenges models to reason beyond observation and predict spatial causal outcomes. We further construct SCP-Bench, a benchmark comprising 2,500 QA pairs across 1,181 videos spanning diverse viewpoints, scenes, and causal directions, to support systematic evaluation. Through comprehensive experiments on {23} state-of-the-art models, we reveal substantial gaps between human and model performance, limited temporal extrapolation, and weak causal grounding. We further analyze key factors influencing performance and propose perception-enhancement and reasoning-guided strategies toward advancing spatial causal intelligence. The project page is https://guangstrip.github.io/SCP-Bench.

Keywords

Cite

@article{arxiv.2603.03944,
  title  = {SCP: Spatial Causal Prediction in Video},
  author = {Yanguang Zhao and Jie Yang and Shengqiong Wu and Shutong Hu and Hongbo Qiu and Yu Wang and Guijia Zhang and Tan Kai Ze and Hao Fei and Chia-Wen Lin and Mong-Li Lee and Wynne Hsu},
  journal= {arXiv preprint arXiv:2603.03944},
  year   = {2026}
}

Comments

30 pages, 21 figures, 17 tables, CVPR findings

R2 v1 2026-07-01T11:02:49.757Z