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

With the rise of pre-trained vision-language models such as CLIP, performing video anomaly detection (VAD) through cross-modal reasoning has become an emerging trend. However, we observe that CLIP still suffers from weak abnormality awareness: normal and abnormal descriptions are highly entangled in the text embedding space, causing video features to assign nearly indistinguishable similarity scores to both types of prompts. To address this issue, we propose Alert-CLIP, an abnormality-aware latentenhanced tuning framework that tailors CLIP for VAD. Alert-CLIP introduces a multi-level alignment strategy:

(1) video-label alignment, which reshapes the semantic space to establish a coarse-level foundation for abnormality awareness; (2) region-text alignment, which explicitly associates anomaly-related regions with detailed descriptions to strengthen fine-grained perception; (3) region-semantic alignment, which contrasts anomalous regions against multiple hard negative samples to enhance abnormality-aware discrimination. To support this training, we construct VAGTA. Extensive experiments show that Alert-CLIP consistently surpasses CLIP across weakly supervised, zeroshot, and open-vocabulary settings. VAGTA is publicly