AI Paper Atlas

Research map / Video / Video segmentation and tracking

TD-VAD: Breaking Visual Dependence in Video Anomaly Detection with Text-Driven Learning

Shuangqing Zhang, Lei-Lei Ma, Zhao Wang, Wen Dong, Xinyi Xu, Guo-Sen Xie, Caifeng Shan, Fang Zhao

ICML 2026 · Video › Video segmentation and tracking · approach: Weakly supervised video anomaly detection · 0 citations (OpenAlex)

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
train VAD on LLM-generated video-like text with frozen CLIP, test on video framesTD-VAD · XD-ViolenceAP (%): 38.21 -> 75.83Improved
add both ECC-Attn and EFC-Attn event-evolution causal attention (ablation)TD-VAD · XD-ViolenceAP (%): 30.48 -> 75.83Improved
add EFC-Attn on top of ECC-Attn (ablation)TD-VAD · XD-ViolenceAP (%): 71.29 -> 75.83Improved
add ECC-Attn on top of EFC-Attn (ablation)TD-VAD · XD-ViolenceAP (%): 73.09 -> 75.83Improved
add binary anomaly head to the multi-class head (hierarchical branch) (ablation)TD-VAD · XD-ViolenceAP (%): 74.83 -> 75.83Improved
add multi-class head to the binary head (ablation)TD-VAD · XD-ViolenceAP (%): 68.22 -> 75.83Improved
increase timestamps per text description from 1 to 4 (ablation)TD-VAD · XD-ViolenceAP (%): 71.79 -> 75.83Improved
increase timestamps from 4 to 8 (ablation)TD-VAD · XD-ViolenceAP (%): 75.83 -> 74.67Hurt
vary EFC-Attn window length (2, 4, 8, 16) (ablation)TD-VAD · XD-ViolenceAP (%): 75.24 / 75.83 / 74.98 / 75.12No clear effect
change LLM prompt style (law enforcement, educational, domain expert) (ablation)text generation · XD-ViolenceAUC change (%): -0.50 / +0.02 / -0.97No clear effect
randomly drop 10-30% of temporal text segments in training (ablation)TD-VAD · XD-ViolenceAP (%): 75.83 -> 76.19 (10% removed)No clear effect
generate training text with Kimi-k2.5 or GPT-3 instead of DeepSeek-V3 (ablation)text generation · XD-ViolenceAP (%): 75.83 -> 74.50 (Kimi-k2.5); 75.54 (GPT-3)Hurt

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