Abstarct: Unusual behavior detection is critically important for visual surveillance. It is also a challenging research topic in computer vision. Although much effort has been devoted to tackle this problem, such detection task in a realistic and uncontrolled environment is still far from mature. The major difficulty lies in the ambiguous characteristic in differentiating normal and abnormal behaviors, whose definitions often vary according to the context of video's history. We propose a frxamework for detecting and locating abnormal activities in video sequences of crowded scenes. The key aspect of our method is the pairing of the spatial-temporal Convolutional Neural Networks (CNN) with handcrafted feature sets such as HOF and HOG for anomaly detection in contiguous video frxames. Handcrafted features learned sparse by using our propose method IW-NMF baxsed on sparse NMF. These feature extracted only from volumes of moving pixels that reduce the computational costs. The architecture of CNN model allow us to extract spatial-temporal features and using handcrafted features to ensure robustness to local noise, and increase detection accuracy. We test our frxamework on popular benchmark datasets containing various human abnormal activities and situations. Evaluation results show that our method outperforms most of other methods and achieves a very competitive detection performance compared to state-of-the-art methods.