Gaussian Mixture Model Algorithm Research Articles

Moving target detection based on improved Gaussian mixture model considering camera motion

This paper proposes a moving target detection scheme suitable for camera motion. Firstly, the background model is initialized by a Gaussian mixture model algorithm. Then Kanade-Lucas-Tomasi Feature Tracker (KLT) method is used to detect optical flow feature points of two adjacent frames, RANdom SAmple Consensus (RANSAC) algorithm is used to filter out the correct matching points and obtain a homography matrix, which can recover the background model matching the current frame, the new background model is used to detect moving target of the current frame. In the foreground detection stage, the current pixel is first compared with its own background model, and then compared with the background model of its 8 neighborhood pixels, the algorithm is speeded up without reducing the detection accuracy in this way; In the update stage of the background model, in order to adapt to the background changes caused by camera motion, an age value variable is set for each pixel. The experimental results show that the improved algorithm has a significant improvement in detection accuracy and running time compared to Gaussian mixture background modeling.

Hierarchical Task-Parameterized Learning from Demonstration for Collaborative Object Movement.

Learning from demonstration (LfD) enables a robot to emulate natural human movement instead of merely executing preprogrammed behaviors. This article presents a hierarchical LfD structure of task-parameterized models for object movement tasks, which are ubiquitous in everyday life and could benefit from robotic support. Our approach uses the task-parameterized Gaussian mixture model (TP-GMM) algorithm to encode sets of demonstrations in separate models that each correspond to a different task situation. The robot then maximizes its expected performance in a new situation by either selecting a good existing model or requesting new demonstrations. Compared to a standard implementation that encodes all demonstrations together for all test situations, the proposed approach offers four advantages. First, a simply defined distance function can be used to estimate test performance by calculating the similarity between a test situation and the existing models. Second, the proposed approach can improve generalization, e.g., better satisfying the demonstrated task constraints and speeding up task execution. Third, because the hierarchical structure encodes each demonstrated situation individually, a wider range of task situations can be modeled in the same framework without deteriorating performance. Last, adding or removing demonstrations incurs low computational load, and thus, the robot's skill library can be built incrementally. We first instantiate the proposed approach in a simulated task to validate these advantages. We then show that the advantages transfer to real hardware for a task where naive participants collaborated with a Willow Garage PR2 robot to move a handheld object. For most tested scenarios, our hierarchical method achieved significantly better task performance and subjective ratings than both a passive model with only gravity compensation and a single TP-GMM encoding all demonstrations.

Quantifying Fracture Networks Inferred From Microseismic Point Clouds by a Gaussian Mixture Model With Physical Constraints

AbstractMicroseismicity is generated by slip on fractures and faults and can be used to infer natural or anthropogenic deformation processes in the subsurface. Yet identifying patterns and fractures from microseismic point clouds is a major challenge that typically relies on the skill and judgment of practitioners. Clustering has previously been applied to tackle this problem, but with limited success. Here, we introduce a probabilistic clustering method to identify fracture networks, based on a Gaussian mixture model algorithm with physical constraints. This method is applied to a rich microseismic data set recorded during the hydraulic fracturing of eight horizontal wells in western Canada. We show that the method is effective for distinguishing hydraulic‐fracture‐created events from induced seismicity. These fractures follow a log‐normal distribution and reflect the physical mechanisms of the hydraulic fracturing process. We conclude that this method has wide applicability for interpreting natural and anthropogenic processes in the subsurface.

A Hybrid Image Segmentation Model Based on GMM and CV Model

Traditional CV level set segmentation model is vulnerable to image noise and non-uniform gray level in the target area, which affects the segmentation accuracy. In this paper, We propose a new image segmentation model, which combines Gaussian mixture model(GMM) and CV level set image segmentation algorithm. We use the GMM foreground detection result as important prior information of CV level set image segmentation in order to integrate multi-information. We add different gray values to foreground area and background area to increase contrast. We put the new different gray values into the level set function to construct new energy terms and it effectively minimize inaccurate edge location caused by the image noise and uneven gray scale in image. In this paper, our method compares to LIF, LBF, RSF and other CV level set image segmentation algorithm, and the experiment result shows that our method is better than others and achieves faster convergence.

An improved Gaussian Mixture Model algorithm for background representation

Initializing a background frame for Gaussian Mixture Model requires no moving objects in the background scene. In this paper, in order to obtain an initial frame when there is a moving object in the background scene, filtering algorithm is used for background frame initialization. This paper proposes an improved method for updating Gaussian mixture models. In the initial stage of the GMM, the update rate of the mean and variance is taken as a larger value, so that the model mean and variance update speed becomes faster, and the model learning speed is accelerated; after training for a period of time with a large update rate, let The mean update rate is unchanged, and the variance update rate becomes smaller, so that the background model can be more stable.

Segmentation of Diabetic Foot Ulcer using Gaussian Mixture Model and Fuzzy KNN Algorithm

In recent years image processing gained the attention of many researches on diabetic foot ulcer (DFU) which is caused by strange increase of blood glucose results nerve injury and leads to amputation. This paper discusses about denoising element on wound by GMM method and image segmentation performed using fuzzy K-NN method. The image is denoised by Gaussian mixture model. The fuzzy KNN algorithm is used for segmentation of wound area from the background image. Further, for analysis the segmented area is subjected to homomorphic filtering and gabor filtering. The performance parameters values such as PSNR, MSE and AMBE are calculated from the resultant image. These values are compared with the Gaussian filtered image values

Improving Accuracy of Evolving GMM Under GPGPU-Friendly Block-Evolutionary Pattern

As a classical clustering model, Gaussian Mixture Model (GMM) can be the footstone of dominant machine learning methods like transfer learning. Evolving GMM is an approximation to the classical GMM under time-critical or memory-critical application scenarios. Such applications often have constraints on time-to-answer or high data volume, and raise high computation demand. A prominent approach to address the demand is GPGPU-powered computing. However, the existing evolving GMM algorithms are confronted with a dilemma between clustering accuracy and parallelism. Point-wise algorithms achieve high accuracy but exhibit limited parallelism due to point-evolutionary pattern. Block-wise algorithms tend to exhibit higher parallelism. Whereas, it is challenging to achieve high accuracy under a block-evolutionary pattern due to the fact that it is difficult to track evolving process of the mixture model in fine granularity. Consequently, the existing block-wise algorithm suffers from significant accuracy degradation, compared to its batch-mode counterpart: the standard EM algorithm. To cope with this dilemma, we focus on the accuracy issue and develop an improved block-evolutionary GMM algorithm for GPGPU-powered computing systems. Our algorithm leverages evolving history of the model to estimate the latest model order in each incremental clustering step. With this model order as a constraint, we can perform similarity test in an elastic manner. Finally, we analyze the evolving history of both mixture components and the data points, and propose our method to merge similar components. Experiments on real images show that our algorithm significantly improves accuracy of the original general purpose bock-wise algorithm. The accuracy of our algorithm is at least comparable to that of the standard EM algorithm and even outperforms the latter under certain scenarios.

Automatic speech patterns recognition of commands using SVM and PSO

This paper proposes the implementation of an Automatic Speech Recognition (ASR) process through extraction of Mel-Frequency Cepstral Coefficients (MFCCs) from voice signal commands, application of the Discrete Cosine Transform (DCT) in these coefficients, Support Vector Machine (SVM) training optimized by the Particle Swarm Optimization (PSO) technique in order to speed up the whole process and using One Against All (OAA) multiclass SVM classification. The main contribution is in training phase that it is the combination of SVM with PSO algorithm, resulting in computational load and processing time reduction. This novel algorithm is called here as PSO-SVM hybrid training application and its performance is shown as the experimental results of voice signal commands in Brazilian Portuguese language. Such commands comprise 10 isolated digits (from zero to nine) and 20 action commands such as “go ahead”, “finish”, “pause”, etc.; that is, there are 30 different pattern types (classes) to be separated (recognized). The process is speaker independent type, that is, the voice bank used in training is different from the one used in tests. The obtained results presented success rates of 92% to 99% during the tests for the classifier using RBF kernel function. Besides, the comparison section shows that this technique is 25 times faster than the recognition without optimization and also, it presents 10% of improvement in recognition success rate when compared to the well-known technique, Gaussian Mixture Models (GMM) algorithm. In addition, the proposed algorithm can be applied in any data processing board for voice signals (DSP, FPGA, DSPIC, ...).

Viewport Prediction for Live 360-Degree Mobile Video Streaming Using User-Content Hybrid Motion Tracking

360-degree video streaming has been gaining popularities recently with the rapid growth of adopting mobile head mounted display (HMD) devices in the consumer video market, especially for live broadcasts. The 360-degree video streaming introduces brand new bandwidth and latency challenges in live streaming due to the significantly increased video data. However, most of the existing bandwidth saving approaches based on viewport prediction have only focused on the video-on-demand (VOD) use cases leveraging historical user behavior data, which is not available in live broadcasts. We develop a new viewport prediction scheme for live 360-degree video streaming using video content-based motion tracking and dynamic user interest modeling. To obtain real-time performance, we implement the Gaussian mixture model (GMM) and optical flow algorithms for motion detection and feature tracking. Then, the user's future viewport of interest is generated by leveraging a dynamic user interest model that weighs all the features and motion information abstracted from the live video frames. Furthermore, we develop two enhancement techniques that take into consideration of user feedback for fast error recovery and view updates. Consequently, our predicted viewports are irregular and dynamically adjusted to cover the maximum portions of the actual user viewports and thus ensure a high prediction accuracy. We evaluate our viewport prediction approach using a public user head movement dataset, which contains the data of 48 users watching 6 360-degree videos. The experimental results show that the proposed approach supports sophisticated user head movement patterns and outperforms the existing velocity-based approach in terms of prediction accuracy. In addition, the motion tracking scheme introduces minimum latency overhead to ensure the quality of live streaming experience.

Adaptive prognosis of centrifugal pump under variable operating conditions

Condition monitoring and prognosis of a centrifugal pump is essential to increase its reliability and safety, and reduce downtime and maintenance costs. Due to the complexity and variety of its operating conditions in practice, the distribution characteristics of data collected from a centrifugal pump shows complex multivariate distributions. On the other hand, the machinery degradation process is also characterized by stochasticity and nonlinearity. To address these challenges, this paper presents an adaptive prognosis method incorporating operating condition identification and Bayesian prediction in one framework. Specifically, a multivariate distribution based unsupervised clustering method is presented to identify operating conditions of pumps based on Gaussian mixture model and Expectation-Maximization algorithm. Then the future machinery status is predicted for defect severity analysis and remaining useful life estimation under the identified operating conditions based on particle filter method. The uncertainty of the prediction results is also quantified and updated with the instrumented measurements. An experimental study on an oil centrifugal pump in the field is performed to demonstrate the effectiveness of the presented adaptive prognosis method over conventional prediction approaches.

Infield oilseed rape images segmentation via improved unsupervised learning models combined with supreme color features

The variability of illumination and weather conditions lead to a big challenge for infield image segmentation. Therefore, robust, fast, and automated algorithms are highly required to obtain reliable image segmentation results. This research was aimed to develop efficient unsupervised clustering algorithms for oilseed rape image segmentation in the field. The Naïve Bayes rule was first employed to select a supreme color feature from ten color models. An initialization approach based on the genetic algorithm (GA) was then used to define the initial cluster centroids for subsequent Gaussian mixture model (GMM), self-organizing map (SOM), fuzzy c-mean (FCM), and k-means algorithms. The length of the chromosome was determined using cluster validity indices. Finally, the performances of these algorithms were evaluated based on the image segmentation quality and computation time. After testing the proposed method on the image datasets from two fields, the results revealed that the highest segmentation accuracy of 96% was obtained using the optimized SOM and the lowest computation time was obtained using the k-means. The GA-based initialization speeded up the convergence process and ensured consistent labeling between runs. All clustering algorithms were proved to be robust to varying illumination conditions and can process images with a very complex background in an automated fashion.

Complex Network Community Extraction Based on Gaussian Mixture Model Algorithm

Based on the problem of community partitioning in complex networks,this paper proposes a Gaussian mixture model community extraction algorithm based on principal component analysis.The idea of the algorithm is as follows:Firstly,the principal component analysis is used to reduce the dimension of the adjacency matrix of the network;secondly,it is assumed that the communities in a network are generated by different Gaussian models,that is,the generation mechanism of different models is different;The parameters of the model are solved by the expectation maximization algorithm. Simulation experiments show that if the contribution rate of the principal component reaches more than 90%, the network division is very consistent with the actual network,and the time used is also short. Compared with other methods,it has obvious advantages.

GPR Imaging for Deeply Buried Objects: A Comparative Study Based on Compositing of Scanning Frequencies and a Chirp Excitation Function

Compositing of ground penetrating radar (GPR) scans of differing frequencies have been found to produce cleaner images at depth using the Gaussian mixture model (GMM) feature of the expectation-maximization (EM) algorithm. GPR scans at various heights (“Stand Off”), as well as ground-based scans, have been studied. In this paper, we compare the GPR response from a chirp excitation function-based radar with the response from the EM GMM algorithm compositing process, using the same mix of frequencies. A chirp excitation pulse was found to be effective in delineating the defined buried object, but the resulting image is less sharp than the GMM EM method.

An EM algorithm for singular Gaussian mixture models

In this paper, we introduce finite mixture models with singular multivariate normal components. These models are useful when the observed data involves collinearities, that is when the covariance matrices are singular. They are also useful when the covariance matrices are ill-conditioned. In the latter case, the classical approaches may lead to numerical instabilities and give inaccurate estimations. Hence, an extension of the Expectation Maximization algorithm, with complete proof, is proposed to derive the maximum likelihood estimators and cluster the data instances for mixtures of singular multivariate normal distributions. The accuracy of the proposed algorithm is then demonstrated on the grounds of several numerical experiments. Finally, we discuss the application of the proposed distribution to financial asset returns modeling and portfolio selection.

Asynchronous Multisonar Data Integration Approach Used to Detect Sparse Targets

Due to the advantage of the distributed multisensor detection system which makes sonar detect further and more accurate estimation of the target state, distributed multisensor data fusion algorithms are widely applied to the sonar detection system. However, on the one hand, the most asynchronous algorithms focus on how to convert asynchronous data fusion into synchronous data fusion. On the other hand, sonar detection system suffers from more serious asynchronous data problems (such as more serious random delay and packet loss defaults) than radar and other fields. Therefore, the traditional asynchronous fusion method has some limitations. When the targets are sparse, this paper proposed a novel asynchronous multisonar data integration approach, in which the Gaussian mixture probability hypothesis density (GMPHD) filter is used to filter clutter for local sonar sensor. Then, the Gaussian mixture model (GMM) algorithm is used to model asynchronous data over a period of time. Finally, all local sonar detection data are integrated into a surveillance region image to help to detect the target. Several simulation tests and a sea test are presented in this paper to test this approach performance.

A hybrid k-means-GMM machine learning technique for turbomachinery condition monitoring

Industrial practise typically applies pre-set original equipment manufacturers (OEMs) limits to turbomachinery online condition monitoring. However, aforementioned technique which considers sensor readings within range as normal state often get overlooked in the developments of degradation process. Thus, turbomachinery application in dire need of a responsive monitoring analysis in order to avoid machine breakdown before leading to a more disastrous event. A feasible machine learning algorithm consists of k-means and Gaussian Mixture Model (GMM) is proposed to observe the existence of signal trend or anomaly over machine active period. The aim of the unsupervised k-means is to determine the number of clusters, k according to the total trend detected from the processed dataset. Next, the designated k is input into the supervised GMM algorithm to initialize the number of components. Experiment results showed that the k-means-GMM model set up not only capable of statistically define machine state conditions, but also yield a time-dependent clustering image in reflecting degradation severity, as a mean to achieve predictive maintenance.

Deep learning based liver cancer detection using watershed transform and Gaussian mixture model techniques

ObjectivesLiver cancer is one of the leading cause of death in all over the world. Detecting the cancer tissue manually is a difficult task and time consuming. Hence, a computer-aided diagnosis (CAD) is used in decision making process for accurate detection for appropriate therapy. Therefore the main objective of this work is to detect the liver cancer accurately using automated method. MethodsIn this work, we have proposed a new system called as watershed Gaussian based deep learning (WGDL) technique for effective delineate the cancer lesion in computed tomography (CT) images of the liver. A total of 225 images were used in this work to develop the proposed model. Initially, the liver was separated using marker controlled watershed segmentation process and finally the cancer affected lesion was segmented using the Gaussian mixture model (GMM) algorithm. After tumor segmentation, various texture features were extracted from the segmented region. These segmented features were fed to deep neural network (DNN) classifier for automated classification of three types of liver cancer i.e. hemangioma (HEM), hepatocellular carcinoma (HCC) and metastatic carcinoma (MET). ResultsWe have achieved a classification accuracy of 99.38%, Jaccard index of 98.18%, at 200 epochs using DNN classifier with a negligible validation loss of 0.062 during the classification process. ConclusionsOur developed system is ready to be tested with huge database and can aid the radiologist in detecting the liver cancer using CT images.

Residual Stresses in Cu/Ni Multilayer Thin Films Measured Using the Sin2ψ Method

Residual stresses in multilayer thin films are of substantial importance to the service life of advanced engineering systems. In this investigation, the residual stresses in magnetron sputtered Cu/Ni multilayer thin films were characterized using x-ray diffraction (XRD) and the sin2ψ method. The influence of layer thickness on residual stress was explored for films with alternating Ni and Cu layers with equal layer thicknesses ranging from 10 nm to 100 nm. To address peak broadening and overlapping, the Gaussian Mixture Model (GMM) and Expectation Maximization (EM) algorithm were employed, and the peak position was determined using the Center of Gravity (CoG) method. Results showed tensile residual stress in both the Cu and Ni layers and a prominent layer thickness dependence. The stress in the Ni layers increased from roughly 880 MPa to 1550 MPa with decreasing layer thickness from 100 nm to 10 nm. In the Cu layers, the stress remained relatively constant at ~250 MPa and then substantially decreased for the 10 nm thickness. The findings confirm that the XRD-based approach can be applied for residual stress measurement in nanoscale multilayer thin films, provided that peak broadening and overlapping issues are addressed. Furthermore, the residual stress in metal multilayers is strongly dependent on layer thickness.

Development of Efficient Strategy for Square Peg-in-Hole Assembly Task

In this paper, we present a dexterous gripper with an angle measurement system and efficient strategy for estimating a hole position from reaction force measurements in a square peg-in-hole task. The contact conditions between a square peg and hole have been analyzed to show that the redundancy in reaction force-moment relation around hole originated from contact conditions. In order to remove such redundancy in contact conditions, an angular error between a square peg and hole is first corrected with the help of the angle measurement system. Then, the positional error between them is identified by applying two clustering algorithms (K means algorithm and Gaussian Mixture Model (GMM) algorithm) to the reaction moment data obtained by the XY search. For grasping and manipulating a peg precisely, a 4-DOF dexterous gripper is designed, which consists of two fingers equipped with 6-axis force/torque sensor at fingertip. The extensive experiments using the proposed gripper with the angle measurement system demonstrate that during the square peg-in-hole task, the proposed strategy ensures the tact time of around 5 sec with the estimation accuracy of less than 0.35 mm and the deviation of less than 0.44 mm, respectively.

Research and Improvement of Content-Based Image Retrieval Framework

This paper proposed a high-performance image retrieval framework, which combines the improved feature extraction algorithm SIFT (Scale Invariant Feature Transform), improved feature matching, improved feature coding Fisher and improved Gaussian Mixture Model (GMM) for image retrieval. Aiming at the problem of slow convergence of traditional GMM algorithm, an improved GMM is proposed. This algorithm initializes the GMM by using on-line [Formula: see text]-means clustering method, which improves the convergence speed of the algorithm. At the same time, when the model is updated, the storage space is saved through the improvement of the criteria for matching rules and generating new Gaussian distributions. Aiming at the problem that the dimension of SIFT (Scale Invariant Feature Transform) algorithm is too high, the matching speed is too slow and the matching rate is low, an improved SIFT algorithm is proposed, which preserves the advantages of SIFT algorithm in fuzzy, compression, rotation and scaling invariance advantages, and improves the matching speed, the correct match rate is increased by an average of 40% to 55%. Experiments on a recently released VOC 2012 database and a database of 20 category objects containing 230,800 images showed that the framework had high precision and recall rates and less query time. Compared with the standard image retrieval framework, the improved image retrieval framework can detect the moving target quickly and effectively and has better robustness.