Abstract:Abnormal event detection in large videos is an important task in research and industrial applications, which has attracted considerable attention in recent years. Existing methods usually solve this problem by extracting local features and then learning an outlier detection model on training videos. However, most previous approaches merely employ hand-crafted visual features, which is a clear disadvantage due to their limited representation capacity. In this paper, we present a novel unsupervised deep feature learning algorithm for the abnormal event detection problem. To exploit the spatiotemporal information of the inputs, we utilize the deep three-dimensional convolutional network (C3D) to perform feature extraction. Then, the key problem is how to train the C3D network without any category labels. Here, we employ the sparse coding results of the hand-crafted features generated from the inputs to guide the unsupervised feature learning. Specifically, we define a multilevel similarity relationship between these inputs according to the statistical information of the shared atoms. In the following, we introduce the quadruplet concept to model the multilevel similarity structure, which could be used to construct a generalized triplet loss for training the C3D network. Furthermore, the C3D network could be utilized to generate the features for sparse coding again, and this pipeline could be iterated for several times. By jointly optimizing between the sparse coding and the unsupervised feature learning, we can obtain robust and rich feature representations. Based on the learned representations, the sparse reconstruction error is applied to predicting the anomaly score of each testing input. Experiments on several publicly available video surveillance datasets in comparison with a number of existing works demonstrate that the proposed algorithm performs favorably against the state-of-the-art methods.

Fast Abnormal Event Detection

Abnormal Event Detection at 150 FPS in MATLAB

Abnormal event detection via the analysis of multi-frame optical flow information

Abnormal Event Detection Based on Analysis of Movement Information of Video Sequence

Video abnormal event detection based on fast sparse coding and surprise computation

A Constrained Self-adaptive Sparse Combination Representation Method for Abnormal Event Detection.

Online Detection of Abnormal Events in Video Streams

Sparse Coding Guided Spatiotemporal Feature Learning for Abnormal Event Detection in Large Videos

Fast Anomaly Detection in Traffic Surveillance Video Based on Robust Sparse Optical Flow

Abnormal Event Detection Via Covariance Matrix for Optical Flow Based Feature.

Fast and accurate detection and localization of abnormal behavior in crowded scenes

Abnormal Events Detection Using Deep Networks for Video Surveillance.

Real-Time Abnormal Event Detection in Complicated Scenes

Abnormal Event Detection Based on Multi-scale Markov Random Field.

An On-Line and Adaptive Method for Detecting Abnormal Events in Videos Using Spatio-Temporal ConvNet

Real-time and Accurate Abnormal Behavior Detection in Videos

Abnormal visual event detection based on multi‐instance learning and autoregressive integrated moving average model in edge‐based Smart City surveillance

Online Least Squares One-Class Support Vector Machines-Based Abnormal Visual Event Detection.

Fast Event Detection on Big Time Series.

Abnormal Global and Local Event Detection in Compressive Sensing Domain

Accelerated Event-Based Feature Detection and Compression for Surveillance Video Systems