AI Paper Atlas

Research map / Video / Video segmentation and tracking

Alert-CLIP: Abnormality-aware Latent-Enhanced Representation Tuning of CLIP for Video Anomaly Detection

Yiyan Zhu, Menghao Zhang, Haifeng Sun, Pengfei Ren, Xianao Chu, Chenye Xu, Hong Tan, Jinghan Wang, Qi Qi, Jingyu Wang

CVPR 2026 · Video › Video segmentation and tracking · approach: Weakly supervised video anomaly detection

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 region alignment loss (ablation)Alert-CLIP global-only · UCF-Crime (weakly supervised)AUC: 88.53 -> 88.85Improved
add semantic hard-negative loss (ablation)Alert-CLIP global+region · UCF-Crime (weakly supervised)AUC: 88.85 -> 89.32Improved
add global alignment only (ablation)CLIP · UCF-Crime (weakly supervised)AUC: 88.02 -> 88.53Improved
two-stage training instead of joint (ablation)Alert-CLIP · UCF-CrimeAUC: 88.61 -> 89.32Improved
hard negatives instead of in-batch negatives (ablation)Alert-CLIP · XD-Violence (zero-shot)AP: 46.44 -> 55.18Improved
hard negatives instead of in-batch negatives (ablation)Alert-CLIP · UCF-Crime (zero-shot)AUC: 72.71 -> 75.75Improved
add stage 2 (region + hard negatives) after stage 1 (ablation)Alert-CLIP · MSAD (weakly supervised)AP: 75.10 -> 78.24Improved
replace CLIP features with Alert-CLIP stage 1VadCLIP · UCF-Crime (weakly supervised)AP: 32.72 -> 32.16No clear effect
replace CLIP with Alert-CLIP (full)zero-shot VAD · UCF-CrimeAUC: 61.91 -> 75.75Improved
replace CLIP with Alert-CLIPImageNet-1K classification · ImageNet-1KTop-1: 68.34 -> 68.58No clear effect
replace CLIP with Alert-CLIPFlickr30k retrieval · Flickr30k I2TR@1: 82.20 -> 89.90Improved
replace OVVAD with Alert-CLIPopen-vocabulary VAD · UCF-CrimeAUC: 86.40 -> 86.14Hurt

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