An overview of deep learning based methods for unsupervised and semi-supervised anomaly detection in videos (2018-01-09T00:00:00.000000Z)

TL;DR

This article reviews the state-of-the-art deep learning based methods for video anomaly detection and categorizes them based on the type of model and criteria of detection.

Abstract

Videos represent the primary source of information for surveillance applications and are available in large amounts but in most cases contain little or no annotation for supervised learning. This article reviews the state-of-the-art deep learning based methods for video anomaly detection and categorizes them based on the type of model and criteria of detection. We also perform simple studies to understand the different approaches and provide the criteria of evaluation for spatio-temporal anomaly detection.

Authors

Dilip Ravi Kiran

1 papers

Mathew Thomas

1 papers

Ranjith Parakkal

1 papers

TL;DR

Abstract

Authors

References87 items

Spatio-Temporal AutoEncoder for Video Anomaly Detection

Abnormal event detection in videos using generative adversarial nets

On the Essence of Unsupervised Detection of Anomalous Motion in Surveillance Videos

Limiting the reconstruction capability of generative neural network using negative learning

Remembering history with convolutional LSTM for anomaly detection

Optical Acceleration for Motion Description in Videos

Training Adversarial Discriminators for Cross-Channel Abnormal Event Detection in Crowds

Energy-Based Localized Anomaly Detection in Video Surveillance

Robust, Deep and Inductive Anomaly Detection

The LV dataset: A realistic surveillance video dataset for abnormal event detection

Deep-Cascade: Cascading 3D Deep Neural Networks for Fast Anomaly Detection and Localization in Crowded Scenes

Spatio-Temporal Anomaly Detection for Industrial Robots through Prediction in Unsupervised Feature Space

Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discovery

Detecting anomalous events in videos by learning deep representations of appearance and motion

Autoencoder with recurrent neural networks for video forgery detection

Abnormal Event Detection in Videos using Spatiotemporal Autoencoder

NIPS 2016 Tutorial: Generative Adversarial Networks

Anomaly Detection in Video Using Predictive Convolutional Long Short-Term Memory Networks

Split-Brain Autoencoders: Unsupervised Learning by Cross-Channel Prediction

Image-to-Image Translation with Conditional Adversarial Networks

STFCN: Spatio-Temporal Fully Convolutional Neural Network for Semantic Segmentation of Street Scenes

Generating Videos with Scene Dynamics

Social LSTM: Human Trajectory Prediction in Crowded Spaces

Video anomaly detection and localisation based on the sparsity and reconstruction error of auto-encoder

Learning Temporal Regularity in Video Sequences

Video anomaly detection using deep incremental slow feature analysis network

Unsupervised Learning of Visual Structure using Predictive Generative Networks

Adversarial Autoencoders

Deep multi-scale video prediction beyond mean square error

Abnormality Detection with Improved Histogram of Oriented Tracklets

Histograms of Optical Flow Orientation and Magnitude to Detect Anomalous Events in Videos

Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting

Efficient sparse-to-dense optical flow estimation using a learned basis and layers

Modeling Representation of Videos for Anomaly Detection using Deep Learning: A Review

Unsupervised Learning of Video Representations using LSTMs

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Crowded Scene Analysis: A Survey

Analyzing Tracklets for the Detection of Abnormal Crowd Behavior

Learning Spatiotemporal Features with 3D Convolutional Networks

Discriminative Autoencoders for Small Targets Detection

Efficient Feature Extraction, Encoding, and Classification for Action Recognition

DL-SFA: Deeply-Learned Slow Feature Analysis for Action Recognition

Large-Scale Video Classification with Convolutional Neural Networks

Detection of Anomalous Crowd Behavior Using Spatio-Temporal Multiresolution Model and Kronecker Sum Decompositions

Stochastic Gradient VB and the Variational Auto-Encoder

Abnormal Event Detection at 150 FPS in MATLAB

Histograms of optical flow orientation for abnormal events detection

Time Varying Autoregressive Moving Average Models for Covariance Estimation

What regularized auto-encoders learn from the data-generating distribution

Video-Based Abnormal Human Behavior Recognition—A Review

Incremental Slow Feature Analysis: Adaptive Low-Complexity Slow Feature Updating from High-Dimensional Input Streams

Representation Learning: A Review and New Perspectives

Video anomaly detection based on local statistical aggregates

Slow Feature Analysis for Human Action Recognition

Sequential Deep Learning for Human Action Recognition

Contractive Auto-Encoders: Explicit Invariance During Feature Extraction

Online detection of unusual events in videos via dynamic sparse coding

Stacked Convolutional Auto-Encoders for Hierarchical Feature Extraction

Anomalous Behaviour Detection Using Spatiotemporal Oriented Energies, Subset Inclusion Histogram Comparison and Event-Driven Processing

Anomaly detection in crowded scenes

Why Does Unsupervised Pre-training Help Deep Learning?

Abnormal events detection based on spatio-temporal co-occurences

Observe locally, infer globally: A space-time MRF for detecting abnormal activities with incremental updates

Extracting and composing robust features with denoising autoencoders

Learning object motion patterns for anomaly detection and improved object detection

Predictive Coding and the Slowness Principle: An Information-Theoretic Approach

Robust Real-Time Unusual Event Detection using Multiple Fixed-Location Monitors

An introduction to ROC analysis

Outlier Detection Using Replicator Neural Networks

Slow Feature Analysis: Unsupervised Learning of Invariances

Mixtures of Probabilistic Principal Component Analyzers

Long Short-Term Memory

Social force model for pedestrian dynamics.

GENERATIVE ADVERSARIAL NETS

Adversarial Autoencoders for Anomalous Event Detection in Images

Cambridge, MA, USA, 2016