AI Paper Atlas

Research map / Video / Video-language understanding

Streaming Video Crime Anticipation with Spatio-Temporal Causal Reasoning

Yusong Wang, Zheyuan Gu, Keyu Mao, Minghao Shao, Mingkun Xu, Prayag Tiwari, Jiawei Shao, Qingsong Zhao

CVPR 2026 · Video › Video-language understanding · approach: Event and anomaly reasoning in video

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add STCRC causal reasoning dataset for training (ablation)STCH + Qwen2-VL-7B · UCF-CrimeWF1: 23.03 -> 40.67Improved
add STCH hypergraph module (ablation)Qwen2-VL-7B · UCF-CrimeWF1: 33.88 -> 40.67Improved
add STCH hypergraph module (ablation)Qwen2-VL-7B · UCF-CrimeTimeDiff: 58.73 -> 55.80Improved
add Group Evolution hyperedges (ablation)STCH · UCF-CrimeWF1: 36.64 -> 40.67Improved
add Causal Cascade Effect hyperedges (ablation)STCH · UCF-CrimeAUC-S: 0.675 -> 0.718Improved
add memory bank for long-stream context (ablation)STCH + Qwen2-VL-7B · UCF-CrimeAUC-L: 0.601 -> 0.692Improved
add STCRC tasks 1 and 2 (ablation)STCH + Qwen2-VL-7B · UCF-CrimeTimeDiff: 58.36 -> 55.80Improved
add STCRC task 5 entity-event causality (ablation)STCH + Qwen2-VL-7B · UCF-CrimeWF1: 29.97 -> 40.67Improved
fine-tune streaming baseline on STCRCFlash-VStream · UCF-CrimeWF1: 23.83 -> 38.07Improved
fine-tune streaming baseline on STCRCVideoLLM-online · UCF-CrimeWF1: 18.89 -> 24.89Improved
use smaller Qwen2-VL-3B backbone instead of 7B (ablation)STCH + STCRC · UCF-CrimeWF1: 40.67 -> 33.15Hurt
use LLaVA-1.5-7B backbone instead of Qwen2-VL-7B (ablation)STCH + STCRC · UCF-CrimeWF1: 40.67 -> 38.27Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy