Multi-granularity Video Unusual Event Detection Based on Infinite Hidden Markov Models
نویسندگان
چکیده
The multi-rate phenomenon of video unusual event is one of the factors to reduce the detection accuracy of video unusual event. Based on the infinite state Hidden Markov Model (iHMM), a multi-granularity detection algorithm for video unusual event is proposed. This algorithm first effectively extracts the feature sequence from the original data through subspace projection technique. Then the feature sequence is sampling at different time intervals to obtain the multi-rate feature sequences. And these multi-rate feature sequences can be used to construct the different time granularities model in the model training stage, and to find the video unusual event at different time granularities in the detection stage. In parameter learning of iHMM, the Beam sampling and EM is combined to improve the efficiency of the iteratively estimation. The experimental results using the surveillance data of vehicles forbidding section, show that the proposed method can be effectively detect unusual events in a complex outdoor scene.
منابع مشابه
Evaluation of the Hidden Markov Model for Detection of P300 in EEG Signals
Introduction: Evoked potentials arisen by stimulating the brain can be utilized as a communication tool between humans and machines. Most brain-computer interface (BCI) systems use the P300 component, which is an evoked potential. In this paper, we evaluate the use of the hidden Markov model (HMM) for detection of P300. Materials and Methods: The wavelet transforms, wavelet-enhanced indepen...
متن کاملSemantic Event Detection in Structured Video Using Hybrid HMM/SVM
In this paper, we propose a new semantic event detection algorithm in structured video. A hybrid method that combines HMM with SVM to detect semantic events in video is proposed. The proposed detection method has some advantages that it is suitable to the temporal structure of event thanks to Hidden Markov Models (HMM) and guarantees high classification accuracy thanks to Support Vector Machine...
متن کاملVideo Event Recognition and Anomaly Detection by Combining Gaussian Process and Hierarchical Dirichlet Process Models
In this paper, we present an unsupervised learning framework for analyzing activities and interactions in surveillance videos. In our framework, three levels of video events are connected by Hierarchical Dirichlet Process (HDP) model: low-level visual features, simple atomic activities, and multi-agent interactions. Atomic activities are represented as distribution of low-level features, while ...
متن کاملLinear Prediction Based Mixture Models for Event Detection in Video Sequences
In this paper, we propose a method for the detection of irregularities in time series, based on linear prediction. We demonstrate how we can estimate the linear predictor by solving the Yule Walker equations, and how we can combine several predictors in a simple mixture model. In several tests, we compare our model to a Gaussian mixture and a hidden Markov model approach. We successfully apply ...
متن کاملEvent-Coupled Hidden Markov Models
Inferences from time-series data can be greatly enhanced by taking into account multiple modalities. In some cases, such as audio of speech and the corresponding video of lip gestures, the different time-series are tightly coupled. We are interested in loosely-coupled time series where only the onset of events are coupled in time. We present an extension of the forward-backward algorithm that c...
متن کامل