Recent advancements in weakly-supervised video anomaly detection have achieved remarkable performance by applying the multiple instance learning paradigm based on multimodal foundation models such as CLIP to highlight anomalous instances and classify categories. However, their objectives may tend to detect the most salient response segments, while neglecting to mine diverse normal patterns separated from anomalies, and are prone to category confusion due to similar appearance, leading to unsatisfactory fine-grained classification results. Therefore, we propose a novel Disentangled Semantic Alignment Network (DSANet) to explicitly separate abnormal and normal features from coarse-grained and fine-grained aspects, enhancing the distinguishability. Specifically, at the coarse-grained level, we introduce a self-guided normality modeling branch that reconstructs input video features under the guidance of learned normal prototypes, encouraging the model to exploit normality cues inherent in the video, thereby improving the temporal separation of normal patterns and anomalous events. At the fine-grained level, we present a decoupled contrastive semantic alignment mechanism, which first temporally decomposes each video into event-centric and background-centric components using frame-level anomaly scores and then applies visual-language contrastive learning to enhance class-discriminative representations. Comprehensive experiments on two standard benchmarks, namely XD-Violence and UCF-Crime, demonstrate that DSANet outperforms existing state-of-the-art methods.

Label distribution learning (LDL) is a novel learning paradigm for solving label ambiguity problems. The existing LDL algorithms are mainly constructed by considering complete supervised information. However, in practical applications, there are often partial missing labels in the label space, which destroys the structure and relevance between labels and makes it difficult to design accurately learning algorithms. To solve this problem, we propose an incomplete label distribution learning method to fill missing labels by exploiting label-specific feature selection. Firstly, we use the sparse learning to obtain the specific features of each class label, and the common features related to all class labels, respectively. Secondly, we apply low rank constraint to obtain the potential label correlations, and the prediction results are corrected by reconstructing the label distribution consisting of non-missing labels. Finally, a linear mapping relationship between features and labels is constructed to recover missing labels. Experimental results on several data sets demonstrate the effectiveness of the proposed method.