Semi-supervised adapted HMMs for unusual event detection

By Dong Zhang, Daniel Gatica-Perez, Samy Bengio, Iain McCowan. Computer Vision and Pattern Recognition, 2005

328 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Signal Processing

View Paper: https://www.semanticscholar.org/paper/417eba8614feacc77212bf46552a0af153a968b2

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 Signal Processing

Citations in the corpus

A Revisit of Sparse Coding Based Anomaly Detection in Stacked RNN Framework (2017, background): “[28] model the normal patterns with a Markov random field (MRF).”
Video-Based Abnormal Human Behavior Recognition—A Review (2012, methodology): “Bayesian adaptation techniques can also be used to create models for unusual events in an unsupervised manner [5], [8],] [96], [122]. proaches have been widely adapted and extended to solve anomaly-detection problems in video-based applications.”
Anomaly Detection and Localization in Crowded Scenes (2014, background): “The authors of [2], [25], and [7] guarantee temporally consistent inference by modeling normal temporal sequences with hidden Markov models (HMMs).”
Remembering history with convolutional LSTM for anomaly detection (2017, background): “All these approaches usually involve two steps, i.e., extract features for appearance or/and motion representation [1], and learn a model in an unsupervised way, such as sparse representation [4], Markov Random Field (MRF) [13], etc., to explore the regular patterns in the training data.”
Cross-dataset action detection (2010, methodology): “Following this approach, a lot of recent work employs GMM to model the temporal structure of audio-visual events [23] or the visual similarities between two images or videos [24].”
Abnormal Event Detection at 150 FPS in MATLAB (2013, methodology): “Graph model normal event representations were proposed in [21, 11, 10, 3, 19, 15, 6, 2], which utilize co-occurrence patterns.”
Modelling Crowd Scenes for Event Detection (2006, background): “Unusual events are analysed in a similar context in [6] and [14] where deviations from example normal behaviour are used to characterise abnormality.”
Future Frame Prediction for Anomaly Detection - A New Baseline (2017, background): “[37] exploit a Markov random filed (MRF) for modeling the normal patterns.”
Deep-anomaly: Fully convolutional neural network for fast anomaly detection in crowded scenes (2016, unknown): “Zhang et al. [30] use a Markov random ﬁeld (MRF) to model the normal patterns of a video with respect to a number of features, such as rarity, unexpectedness, and relevance.”
SIFT-Bag kernel for video event analysis (2008, methodology): “Consequently, each video clip is represented by a specialized GMM distribution model, and a natural similarity measure between them is the Kullback-Leibler divergence, ga(z) D(ga||gb)=ga(z)logdz.”
Video Behavior Profiling for Anomaly Detection (2008, methodology): “A semisupervised model was introduced in [33] with a two-stage training process.”