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

By Jaechul Kim, Kristen Grauman. 2009

801 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

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 Vision and Pattern Recognition

Citations in the corpus

C3D: Generic Features for Video Analysis (2014, background): “Over the years various problems like action recognition [2, 31, 40], anomaly detection [24], video retrieval [1] and many more have been proposed.”
Anomaly Detection and Localization in Crowded Scenes (2014, methodology): “These include the modeling of motion patterns with histograms of pixel change [5], histograms of optical flow [19], [8], [20], or optical flow measures [3], [4], [17], [1].”
Latent Space Autoregression for Novelty Detection (2018, methodology): “In this respect, prior works involve handcrafted features such as optical flow or trajectory analysis and, on top of that, employ both non-parametric [1] and parametric [5, 28, 22] estimators, as well as graphical modeling [16, 20].”
AnomalyNet: An Anomaly Detection Network for Video Surveillance (2019, background): “For example, histogram of oriented gradients (HOG) [8], 3D spatiotemporal gradient [9], and the histogram of oriented flows (HOF) [10] have been extensively used in [6] and [11]–[13].”
Object-Centric Auto-Encoders and Dummy Anomalies for Abnormal Event Detection in Video (2018, methodology): “Most existing anomaly detection approaches [2, 5, 15, 18, 23, 25, 26, 37, 39] are based on outlier detection and learn a model of normality from training videos containing only familiar events.”
Video anomaly detection based on local statistical aggregates (2012, background): “This insight helps collapse the large ambient data dimension for detecting local anomalies.”
Deep-anomaly: Fully convolutional neural network for fast anomaly detection in crowded scenes (2016, neutral): “Kim and Grauman [35] A method based on motion properties of pixels for behavior modeling is proposed by Benezeth et al. [36].”
Learning Deep Representations of Appearance and Motion for Anomalous Event Detection (2015, methodology): “[11] propose a method based on local optical ﬂow features and MRFs to spot unusual events.”
Abnormal event detection in crowded scenes using sparse representation (2013, methodology): “Kim et al.[16] model local optical flow with MPPCA and enforce consistency by Markov Random Field.”
Abnormal Event Detection in Videos using Spatiotemporal Autoencoder (2017, neutral): “Space-time MRF [8] N/A 56/3 18/0 Online [4] N/A 60/5 19/2 ConvAE [5] 45/4 61/15 17/5 Ours 44/6 61/9 18/10”
Adversarially Learned One-Class Classifier for Novelty Detection (2018, background): “IBC [6] 13% RE [36] 15% MPCCA [19] 30% Ravanbakhsh et al.”
Memorizing Normality to Detect Anomaly: Memory-Augmented Deep Autoencoder for Unsupervised Anomaly Detection (2019, methodology): “Kim and Grauman [15] use a mixture of probabilistic PCA (MPPCA) to model optical flow features.”
Abnormal event detection in videos using generative adversarial nets (2017, methodology): “However, despite a lot of research has been done in this area in the past years [1, 2, 3, 4, 5, 6, 7], the problem is still open.”
Abnormal Event Detection at 150 FPS in MATLAB (2013, methodology): “Our results are comparable to those of [22, 10, 5].”
Anomaly Detection in Video Sequence With Appearance-Motion Correspondence (2019, methodology): “We performed the evaluation according to the ground truth of events with the training and test sets provided in [20], in which the normal events in the ﬁrst 15 minutes of the Entrance Gate video and 5 minutes of the Exit Gate were used in training stage.”
A Revisit of Sparse Coding Based Anomaly Detection in Stacked RNN Framework (2017, methodology): “To represent the local optical flow patterns, Kim and Grauman [14] utilize a mixture of probabilistic PCA models.”
Online detection of unusual events in videos via dynamic sparse coding (2011, methodology): “The fist 5 minutes of the video, same as in [10], is used to build initial dictionary.”