Abnormal Event Detection in Videos using Spatiotemporal Autoencoder

By Yong Shean Chong, Yong Haur Tay. International Symposium on Neural Networks, 2017

596 citations recorded by the source corpus.

We present an efficient method for detecting anomalies in videos. Recent applications of convolutional neural networks have shown promises of convolutional layers for object detection and recognition, especially in images. However, convolutional neural networks are supervised and require labels as learning signals. We propose a spatiotemporal architecture for anomaly detection in videos including crowded scenes. Our architecture includes two main components, one for spatial feature representation, and one for learning the temporal evolution of the spatial features. Experimental results on Avenue, Subway and UCSD benchmarks confirm that the detection accuracy of our method is comparable to state-of-the-art methods at a considerable speed of up to 140 fps.

Subfields: Artificial Intelligence, Computer Networks and Communications, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/527cc8cd2af06a9ac2e5cded806bab5c3faad9cf

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Networks and Communications

In Computer Vision and Pattern Recognition

Citations in the corpus

Learning Regularity in Skeleton Trajectories for Anomaly Detection in Videos (2019, methodology): “These features are usually extracted from whole frames [5, 13, 18, 20, 25], localized on a grid of image patches [24], or concentrated on pre-identified regions [6, 14].”
GANomaly: Semi-Supervised Anomaly Detection via Adversarial Training (2018, background): “Much of the recent work on this problem domain pays particular attention to the use of spatiotemporal features by modeling spatiotemporal CNN [22], 3D convolutional autoencoders [23, 24] and long short-term memories [25].”
Memorizing Normality to Detect Anomaly: Memory-Augmented Deep Autoencoder for Unsupervised Anomaly Detection (2019, background): “Some very recent works [43, 46, 48, 6] train deep autoencoders for anomaly detection.”
Future Frame Prediction for Anomaly Detection - A New Baseline (2017, methodology): “Thus, by taking both advantages of CNN and RNN, [5][21] leverage a Convolutional LSTMs Auto-Encoder (ConvLSTM-AE) to model normal appearance and motion patterns at the same time, which further boosts the performance of the Conv-AE based solution.”
Graph Convolutional Label Noise Cleaner: Train a Plug-And-Play Action Classifier for Anomaly Detection (2019, background): “Recently, with the great success of deep learning, a few researchers design deep neural networks on abstraction feature learning [19, 12, 42] or video prediction learning [40] for anomaly detection.”
An overview of deep learning based methods for unsupervised and semi-supervised anomaly detection in videos (2018, methodology): “The spatio-temporal au-toencoder in [42] is shown in the right panel of ﬁgure 4.”
Anomaly Detection using One-Class Neural Networks (2018, background): “Following the unprecedented success of using deep autoencoder networks, as feature extractors, in tasks as diverse as visual, speech anomaly detection [13, 27], several hybrid models that combine feature extraction using deep learning and OC-SVM have appeared [16, 32].”
Deep Autoencoding Models for Unsupervised Anomaly Segmentation in Brain MR Images (2018, background): “Similar to [5,2], we then make the architecture fully convolutional to ensure that spatial information is not lost in the bottleneck of the model.”
Learning Memory-Guided Normality for Anomaly Detection (2020, methodology): “Many anomaly detection methods leverage reconstructive models [9, 26, 5, 33] exploiting feature representations from e.”
AnomalyNet: An Anomaly Detection Network for Video Surveillance (2019, background): “For example, [14] and [15] proposed a neural network which consists of a recurrent”
A Revisit of Sparse Coding Based Anomaly Detection in Stacked RNN Framework (2017, methodology): “Following the work [11][3], after calculating all frame level reconstruction errors for all testing videos, we normalize the errors to range [0, 1], and calculate regularity score for each frame based on the following equation:”