In the paper we propose a face verifying algorithm for face recognition that can identify two face mismatch pairs in cases of incorrect decisions. The computational approach taken in this system is performed by the de...
详细信息
In the paper we propose a face verifying algorithm for face recognition that can identify two face mismatch pairs in cases of incorrect decisions. The computational approach taken in this system is performed by the derivative of accumulated absolute difference between two faces unseen before. Unlike the traditional multi-dimensional distance measurement, the proposed algorithm also considers an increasing trend of accumulated absolute difference in respect to the Gaussian components. A Gaussian mixture model of bag-of-feature from training faces is also widely applicable to several biometric systems. Evaluation of the proposed algorithm is done on unconstrained environments using Labeled Face in the Wild (LFW) datasets. Experiments show that the proposed algorithm outperforms all conventional face recognition algorithms with advantage of about 4.92% over direct-bag-of-features and 18.05% over principal component analysis-based and is also appropriate for identification task of the face recognition systems. Furthermore, some particular advantages of our approach are that it can be applied to other verification systems. (C) 2016 The Authors. Published by Elsevier B.V.
The performance of traditional classification models can adversely be impacted by the presence of label noise in training observations. The pioneer work of Lawrence and Scholkopf tackled this issue in datasets with in...
详细信息
ISBN:
(纸本)9783642237805
The performance of traditional classification models can adversely be impacted by the presence of label noise in training observations. The pioneer work of Lawrence and Scholkopf tackled this issue in datasets with independent observations by incorporating a statistical noise model within the inference algorithm. In this paper, the specific case of label noise in non-independent observations is rather considered. For this purpose, a label noise-tolerant expectation-maximisationalgorithm is proposed in the frame of hidden Markov models. Experiments are carried on both healthy and pathological electrocardiogram signals with distinct types of additional artificial label noise. Results show that the proposed label noise-tolerant inference algorithm can improve the segmentation performances in the presence of label noise.
In this article we consider HMM parameter estimation in the the context of an EM algorithm. The models we study are discrete time Markov chains observed in Gaussian noise. Now formulae for up-dating smoothed estimates...
详细信息
ISBN:
(纸本)0780370619
In this article we consider HMM parameter estimation in the the context of an EM algorithm. The models we study are discrete time Markov chains observed in Gaussian noise. Now formulae for up-dating smoothed estimates are given for the models just described. These formulae are computed by exploiting a duality between a forward in time unnormalised probability process and its dual.
Vehicle motion models are employed in driver assistance systems for tracking and prediction tasks. For probabilistic decision making and uncertainty propagation, the prediction's inaccuracy is taken into account i...
详细信息
ISBN:
(纸本)9781467365963
Vehicle motion models are employed in driver assistance systems for tracking and prediction tasks. For probabilistic decision making and uncertainty propagation, the prediction's inaccuracy is taken into account in the form of process noise. This work estimates Gaussian process noise models from measured vehicle trajectories using the expectationmaximisation (EM) algorithm. The method is exemplified and the results evaluated for three commonly used motion models based on a large-scale dataset. A novel closed-form adaptation of the algorithm to a covariance matrix with Kronecker product structure, as in models for translational motion, is presented. The findings suggest that the longitudinal prediction errors feature a non-Gaussian distribution but a reasonable approximation is given by the estimated model.
In recent years, the progress of the Internet of Things has promoted data utilisation in manufacturing industries and has created new possibilities for monitoring the condition of production equipment. By applying ano...
详细信息
In recent years, the progress of the Internet of Things has promoted data utilisation in manufacturing industries and has created new possibilities for monitoring the condition of production equipment. By applying anomaly detection procedures to the data acquired from sensors, it is possible to capture early signs of occurring anomalies, which leads to improvement of operating rates and prevention of accidents. However, in conventional anomaly detection procedures, it is not always possible to properly detect anomalies when usage situations change. This is because the definition of anomalies changes depending on the usage situation. In other words, when 'environment variables' indicating usage conditions and 'monitoring variables' indicating monitoring targets exist, it is necessary to regard them as a conditional anomaly detection problem, which is a problem of detecting anomalies occurring in a monitoring variable on the condition that an environmental variable has occurred. In this paper, we propose a novel analysis procedure to solve such conditional anomaly detection problems. In particular, we propose a conditional anomaly detection procedure when categorical environmental variables and continuous monitoring variables are observed. Through Monte Carlo simulation, we show that the proposed procedure can accurately detect 'conditional anomalies' that cannot be detected by conventional procedures.
Autoregressive Markov switching (ARMS) time series models are used to represent real-world signals whose dynamics may change over time. They have found application in many areas of the natural and social sciences, as ...
详细信息
Autoregressive Markov switching (ARMS) time series models are used to represent real-world signals whose dynamics may change over time. They have found application in many areas of the natural and social sciences, as well as in engineering. In general, inference in this kind of systems involves two problems: (a) detecting the number of distinct dynamical models that the signal may adopt and (b) estimating any unknown parameters in these models. In this paper, we introduce a new class of nonlinear ARMS time series models with delays that includes, among others, many systems resulting from the discretisation of stochastic delay differential equations (DDEs). Remarkably, this class includes cases in which the discretisation time grid is not necessarily aligned with the delays of the DDE, resulting in discrete-time ARMS models with real (non-integer) delays. The incorporation of real, possibly long, delays is a key departure compared to typical ARMS models in the literature. We describe methods for the maximum likelihood detection of the number of dynamical modes and the estimation of unknown parameters (including the possibly non-integer delays) and illustrate their application with a nonlinear ARMS model of El Ni & ntilde;o-southern oscillation (ENSO) phenomenon.
Text clustering is an important method for effectively organising, summarising, and navigating text information. However, in the absence of labels, the text data to be clustered cannot be used to train the text repres...
详细信息
Text clustering is an important method for effectively organising, summarising, and navigating text information. However, in the absence of labels, the text data to be clustered cannot be used to train the text representation model based on deep learning. To address the problem, an algorithm of text clustering based on deep representation learning is proposed using the transfer learning domain adaptation and the parameters update during cluster iteration. First, source domain data is used to perform the pre-training of the deep learning classification model. This procedure acts as an initialisation of the model parameters. Then, the domain discriminator is added to the model, to domain-divide the input sample. If the discriminator cannot distinguish which domain the data belongs to, the common feature space of two domains is obtained, so the domain adaptation problem is solved. Finally, the text feature vectors obtained by the model are clustered with MCSKM++ algorithm. The algorithm not only resolves the model pre-training problem in unsupervised clustering, but also has a good clustering effect on the transfer problem caused by different numbers of domain labels. Experiments suggest that the clustering accuracy of the algorithm is superior to other similar algorithms.
This paper is concerned with the parameter estimation of a relatively general class of nonlinear dynamic systems. A Maximum Likelihood (ML) framework is employed in the interests of statistical efficiency, and it is i...
详细信息
This paper is concerned with the parameter estimation of a relatively general class of nonlinear dynamic systems. A Maximum Likelihood (ML) framework is employed in the interests of statistical efficiency, and it is illustrated how an expectationmaximisation (EM) algorithm may be used to compute these ML estimates. An essential ingredient is the employment of so-called “particle smoothing„ methods to compute required conditional expectations via a Monte Carlo approach. A simulation example demonstrates the efficacy of these techniques.
Despite developments in sensor technology, monitoring a biological process using regular sensor measurements is often difficult. Development of Bayesian state observers, such as extended Kalman filter(EKF), is an attr...
详细信息
Despite developments in sensor technology, monitoring a biological process using regular sensor measurements is often difficult. Development of Bayesian state observers, such as extended Kalman filter(EKF), is an attractive alternative for soft-sensing of such complex systems. The performance of EKF is dependent on the accurate characterisation of the uncertainties in the state dynamics and in the measurements. In this work, an extended expectationmaximisation (EM) algorithm is developed for estimation of the state and measurement noise covariances for the EKF using irregularly sampled multi-rate measurements. The efficacy of the proposed approach is demonstrated on a benchmark continuous fermenter system. The simulation results reveal that the proposed approach generates fairly accurate estimates of the noise covariances.
In the paper we propose a face verifying algorithm for face recognition that can identify two face mismatch pairs in cases of incorrect decisions. The computational approach taken in this system is performed by the de...
详细信息
In the paper we propose a face verifying algorithm for face recognition that can identify two face mismatch pairs in cases of incorrect decisions. The computational approach taken in this system is performed by the derivative of accumulated absolute difference between two faces unseen before. Unlike the traditional multi-dimensional distance measurement, the proposed algorithm also considers an increasing trend of accumulated absolute difference in respect to the Gaussian components. A Gaussian mixture model of bag-of-feature from training faces is also widely applicable to several biometric systems. Evaluation of the proposed algorithm is done on unconstrained environments using Labeled Face in the Wild (LFW) datasets. Experiments show that the proposed algorithm outperforms all conventional face recognition algorithms with advantage of about 4.92% over direct-bag-of-features and 18.05% over principal component analysis-based and is also appropriate for identification task of the face recognition systems. Furthermore, some particular advantages of our approach are that it can be applied to other verification systems.
暂无评论