Chih-Yung Chang

Wen-Dong Jiang, Chih-Yung Chang, Diptendu Sinha Roy

Recently, violence detection systems developed using unified multimodal models have achieved significant success and attracted widespread attention. However, most of these systems face two critical challenges: the lack of interpretability as black-box models and limited functionality, offering only classification or retrieval capabilities. To address these challenges, this paper proposes a novel interpretable violence detection system, termed the Three-in-One (TIO) System. The TIO system integrates knowledge graphs (KG) and graph attention networks (GAT) to provide three core functionalities: detection, retrieval, and explanation. Specifically, the system processes each video frame along with text descriptions generated by a large language model (LLM) for videos containing potential violent behavior. It employs ImageBind to generate high-dimensional embeddings for constructing a knowledge graph, uses GAT for reasoning, and applies lightweight time series modules to extract video embedding features. The final step connects a classifier and retriever for multi-functional outputs. The interpretability of KG enables the system to verify the reasoning process behind each output. Additionally, the paper introduces several lightweight methods to reduce the resource consumption of the TIO system and enhance its efficiency. Extensive experiments conducted on the XD-Violence and UCF-Crime datasets validate the effectiveness of the proposed system. A case study further reveals an intriguing phenomenon: as the number of bystanders increases, the occurrence of violent behavior tends to decrease.

Wen-Dong Jiang, Chih-Yung Chang, Show-Jane Yen et al.

Deep learning has achieved remarkable success in processing and managing unstructured data. However, its "black box" nature imposes significant limitations, particularly in sensitive application domains. While existing interpretable machine learning methods address some of these issues, they often fail to adequately consider feature correlations and provide insufficient evaluation of model decision paths. To overcome these challenges, this paper introduces Real Explainer (RealExp), an interpretability computation method that decouples the Shapley Value into individual feature importance and feature correlation importance. By incorporating feature similarity computations, RealExp enhances interpretability by precisely quantifying both individual feature contributions and their interactions, leading to more reliable and nuanced explanations. Additionally, this paper proposes a novel interpretability evaluation criterion focused on elucidating the decision paths of deep learning models, going beyond traditional accuracy-based metrics. Experimental validations on two unstructured data tasks -- image classification and text sentiment analysis -- demonstrate that RealExp significantly outperforms existing methods in interpretability. Case studies further illustrate its practical value: in image classification, RealExp aids in selecting suitable pre-trained models for specific tasks from an interpretability perspective; in text classification, it enables the optimization of models and approximates the performance of a fine-tuned GPT-Ada model using traditional bag-of-words approaches.

Wen-Dong Jiang, Chih-Yung Chang, Hsiang-Chuan Chang et al.

Weakly Supervised Monitoring Anomaly Detection (WSMAD) utilizes weak supervision learning to identify anomalies, a critical task for smart city monitoring. However, existing multimodal approaches often fail to meet the real-time and interpretability requirements of edge devices due to their complexity. This paper presents TCVADS (Two-stage Cross-modal Video Anomaly Detection System), which leverages knowledge distillation and cross-modal contrastive learning to enable efficient, accurate, and interpretable anomaly detection on edge devices.TCVADS operates in two stages: coarse-grained rapid classification and fine-grained detailed analysis. In the first stage, TCVADS extracts features from video frames and inputs them into a time series analysis module, which acts as the teacher model. Insights are then transferred via knowledge distillation to a simplified convolutional network (student model) for binary classification. Upon detecting an anomaly, the second stage is triggered, employing a fine-grained multi-class classification model. This stage uses CLIP for cross-modal contrastive learning with text and images, enhancing interpretability and achieving refined classification through specially designed triplet textual relationships. Experimental results demonstrate that TCVADS significantly outperforms existing methods in model performance, detection efficiency, and interpretability, offering valuable contributions to smart city monitoring applications.

Wen-Dong Jiang, Chih-Yung Chang, Ssu-Chi Kuai et al.

The growing demand for intelligent security in consumer electronics, such as smart home cameras and personal monitoring systems, is often hindered by the high computational cost and large model sizes of advanced AI. These limitations prevent the effective deployment of real-time Video Anomaly Detection (VAD) on resource-constrained edge devices. To bridge this gap, this paper introduces Rule-based Video Anomaly Detection (RuleVAD), a novel, lightweight system engineered for high-efficiency and low-complexity threat detection directly on consumer hardware. RuleVAD features an innovative decoupled dual-branch architecture to minimize computational load. An implicit branch uses visual features for rapid, coarse-grained binary classification, efficiently filtering out normal activity to avoid unnecessary processing. For potentially anomalous or complex events, a multimodal explicit branch takes over. This branch leverages YOLO-World to detect objects and applies data mining to generate interpretable, text-based association rules from the scene. By aligning these rules with visual data, RuleVAD achieves a more nuanced, fine-grained classification, significantly reducing the false alarms common in vision-only systems. Extensive experiments on the XD-Violence and UCF-Crime benchmark datasets show that RuleVAD achieves superior performance, surpassing existing state-of-the-art methods in both accuracy and speed. Crucially, the entire system is optimized for low-power operation and is fully deployable on an NVIDIA Jetson Nano board, demonstrating its practical feasibility for bringing advanced, real-time security monitoring to everyday consumer electronic devices.