Gaussian Mixed Model Research Articles

Video SAR Moving Target Detection Method Based on Machine Vision

When detecting moving targets in video SAR, it is impossible to accurately obtain the moving targets in dynamic images. In order to solve the problems of low detection integrity and poor effectiveness, a video SAR moving target detection method based on machine vision is proposed. Firstly, the moving target pixel points are matched by the mixed Gaussian model to obtain the target area of the video SAR moving image; secondly, the HSV model and reflectivity algorithm are used to remove the shadows caused by the moving target; finally, the processed target area is input into the YOLO algorithm to complete the final video SAR moving target detection. Experimental results show that the proposed algorithm has high detection integrity, high detection rate, low false detection rate, and better detection effectiveness.

A Novel Fault Feature Selection and Diagnosis Method for Rotating Machinery with SI-IR2CMSE and SSGMM-SR

Abstract In the field of fault diagnosis, machine learning is highly valued for its broad applicability and efficiency. Feature extraction and feature selection are key steps in the application of machine learning, and the performance of fault diagnosis methods relies heavily on the effective execution of these two steps. For this reason, this paper aims to enhance the performance of fault diagnosis methods by improving these two aspects. Firstly, to address the non-linearity and non-stationarity of rotating machinery vibration signals under variable operation conditions, this paper proposes an improved rapid refined composite multiscale sample entropy (IR2CMSE) feature extraction method. In addition, this paper decomposes the vibration signals with improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) and extracts the sensitive intrinsic modal functions’ (IMFs') IR2CMSE values (SI-IR2CMSE) as the initial feature vector, which more accurately reveals the intrinsic time-scale characteristics of the vibration signals. Secondly, to address the problem of over-reliance on sample labels in most feature selection methods, this paper proposes a semi-supervised Gaussian mixing model with sparse regularization (SSGMM-SR) feature selection model. The model does not require complete fault labels and can automatically identify important features. Finally, validation with two rotating machinery fault datasets shows that the method proposed in this study exhibits high diagnostic accuracy and stability across multiple classifiers.

Underground concrete pipe crack damage monitoring by the fuzzy analysis of microseism

Abstract Underground concrete pipe is critical for the safety of the urban infrastructure. Crack would lead to the accidents such as fluid spillage, ground subsidence, as well as waterlogging. Thus, it is of great significance to measure and monitor the crack of pipeline. In this research, a fuzzy monitoring method of the concrete pipe crack damage based on the microseism signal GMM (Gaussian Mixed Model) analysis was proposed. First, the multi parameters of microseism signal were extracted. Then, CRITIC (Criteria Importance Through Intercriteria Correlation) weight vector was constructed using the importance information of parameters. Further, the GMM membership matrix was created by the similarity of the probability density distribution. The crack condition was assessed by means of the fuzzy calculation between the weight vector and membership matrix. The experiment indicated that the proposed method could monitor and assess the different crack conditions in real time.

Multiscale experimental characterization and physical-based constitutive modeling of silicone adhesive

Silicone adhesive has been widely used in assembly glass curtain walls which leads to its mechanical behaviors under heavy investigation. Due to its complexity, the microstructure of silicone adhesive stays unclear. A detailed description of microstructures of silicone adhesive is the basis to obtain an in-depth understanding of its mechanical behavior and the related mechanisms. In this work, series of multiscale experiments, uniaxial tension, uniaxial compression, simple shear and scanning electron microscopy (SEM) observation, were conducted to characterize mechanical behaviors and microstructure features of silicone adhesive. Stress-strain curves obtained through macroscale experiments were presented. Morphology and dispersion of filler clusters in silicone adhesive were analyzed using Gaussian mixed models. A modification of molecular chain spatial distribution was made to non-affine network model to consider anisotropic damage evolution of polymer matrix. A simplified term was further introduced to account for load-bearing capacity of filler clusters based on fractal theory where fractal dimensions of clusters work as a connection between microscale and macroscale. The modified model was further extended to describe Mullins effect combined with network alteration theory. The results of model validation and comparison demonstrated that due to the consideration of polymer chain anisotropic evolution and filler cluster microscopic features, the proposed model can predict the mechanical behavior of silicone adhesive under different loading conditions as well as the stress softening and permanent set of Mullins effect more accurately.

G-T correcting: an improved training of image segmentation under noisy labels.

Data-driven medical image segmentation networks require expert annotations, which are hard to obtain. Non-expert annotations are often used instead, but these can be inaccurate (referred to as "noisy labels"), misleading the network's training and causing a decline in segmentation performance. In this study, we focus on improving the segmentation performance of neural networks when trained with noisy annotations. Specifically, we propose a two-stage framework named "G-T correcting," consisting of "G" stage for recognizing noisy labels and "T" stage for correcting noisy labels. In the "G" stage, a positive feedback method is proposed to automatically recognize noisy samples, using a Gaussian mixed model to classify clean and noisy labels through the per-sample loss histogram. In the "T" stage, a confident correcting strategy and early learning strategy are adopted to allow the segmentation network to receive productive guidance from noisy labels. Experiments on simulated and real-world noisy labels show that this method can achieve over 90% accuracy in recognizing noisy labels, and improve the network's DICE coefficient to 91%. The results demonstrate that the proposed method can enhance the segmentation performance of the network when trained with noisy labels, indicating good clinical application prospects.

A machine learning and drug repurposing approach to target ferroptosis in colorectal cancer stratified by sex and KRAS.

The landscape of sex differences in Colorectal Cancer (CRC) has not been well characterized with respect to the mechanisms of action for oncogenes such as KRAS. However, our recent study showed that tumors from male patients with KRAS mutations have decreased iron-dependent cell death called ferroptosis. Building on these findings, we further examined ferroptosis in CRC, considering both sex of the patient and KRAS mutations, using public databases and our in-house CRC tumor cohort. Through subsampling inference and variable importance analysis (VIMP), we identified significant differences in gene expression between KRAS mutant and wild type tumors from male patients. These genes suppress (e.g., SLC7A11 ) or drive (e.g., SLC1A5 ) ferroptosis, and these findings were further validated with Gaussian mixed models. Furthermore, we explored the prognostic value of ferroptosis regulating genes and discovered sex- and KRAS-specific differences at both the transcriptional and metabolic levels by random survival forest with backward elimination algorithm (RSF-BE). Of note, genes and metabolites involved in arginine synthesis and glutathione metabolism were uniquely associated with prognosis in tumors from males with KRAS mutations. Additionally, drug repurposing is becoming popular due to the high costs, attrition rates, and slow pace of new drug development, offering a way to treat common and rare diseases more efficiently. Furthermore, increasing evidence has shown that ferroptosis inhibition or induction can improve drug sensitivity or overcome chemotherapy drug resistance. Therefore, we investigated the correlation between gene expression, metabolite levels, and drug sensitivity across all CRC primary tumor cell lines using data from the Genomics of Drug Sensitivity in Cancer (GDSC) resource. We observed that ferroptosis suppressor genes such as DHODH , GCH1 , and AIFM2 in KRAS mutant CRC cell lines were resistant to cisplatin and paclitaxel, underscoring why these drugs are not effective for these patients. The comprehensive map generated here provides valuable biological insights for future investigations, and the findings are supported by rigorous analysis of large-scale publicly available data and our in-house cohort. The study also emphasizes the potential application of VIMP, Gaussian mixed models, and RSF-BE models in the multi-omics research community. In conclusion, this comprehensive approach opens doors for leveraging precision molecular feature analysis and drug repurposing possibilities in KRAS mutant CRC.

Strategies to determine positive anti-SARS-CoV-2 memory T lymphocyte response during the evolution of an epidemic

During SARS-CoV-2 pandemic, the assessment of immune protection of people at risk of severe infection was an important goal. The appearance of VOCs (Variant of Concern) highlighted the limits of evaluating immune protection through the humoral response. While the humoral response partly loses its neutralizing activity, the anti-SARS-CoV-2 memory T cell response strongly cross protects against VOCs becoming an indispensable tool to assess immune protection. We compared two techniques available in laboratory to evaluate anti-SARS-CoV-2 memory T cell response in a cohort of infected or vaccinated patients with different levels of risk to develop a severe disease: the ELISpot assay and the T-Cell Lymphocyte Proliferation Assay respectively exploring IFNγ production and cell proliferation. We showed that the ELISpot assay detected more anti-Spike memory T cell response than the Lymphocyte Proliferation Assay. We next observed that the use of two different suppliers as antigenic source in the ELISpot assay did not affect the detection of anti-Spike memory T cell response. Finally, we explored a new approach for defining the positivity threshold, using unsupervised mixed Gaussian modeling, challenging the traditional ROC curve used by the supplier. That will be helpful in endemic situation where it could be difficult to recruit “negative” patients.

Outcome-guided Bayesian clustering for disease subtype discovery using high-dimensional transcriptomic data

Due to the tremendous heterogeneity of disease manifestations, many complex diseases that were once thought to be single diseases are now considered to have disease subtypes. Disease subtyping analysis, that is the identification of subgroups of patients with similar characteristics, is the first step to accomplish precision medicine. With the advancement of high-throughput technologies, omics data offers unprecedented opportunity to reveal disease subtypes. As a result, unsupervised clustering analysis has been widely used for this purpose. Though promising, the subtypes obtained from traditional quantitative approaches may not always be clinically meaningful (i.e. correlate with clinical outcomes). On the other hand, the collection of rich clinical data in modern epidemiology studies has the great potential to facilitate the disease subtyping process via omics data and to discovery clinically meaningful disease subtypes. Thus, we developed an outcome-guided Bayesian clustering (GuidedBayesianClustering) method to fully integrate the clinical data and the high-dimensional omics data. A Gaussian mixed model framework was applied to perform sample clustering; a spike-and-slab prior was utilized to perform gene selection; a mixture model prior was employed to incorporate the guidance from a clinical outcome variable; and a decision framework was adopted to infer the false discovery rate of the selected genes. We deployed conjugate priors to facilitate efficient Gibbs sampling. Our proposed full Bayesian method is capable of simultaneously (i) obtaining sample clustering (disease subtype discovery); (ii) performing feature selection (select genes related to the disease subtype); and (iii) utilizing clinical outcome variable to guide the disease subtype discovery. The superior performance of the GuidedBayesianClustering was demonstrated through simulations and applications of breast cancer expression data and Alzheimer's disease. An R package has been made publicly available on GitHub to improve the applicability of our method.

221 Impact of parity on stall use in free access stall systems

Abstract Consumer pressure is leading producers to transition to gestational group-housing systems. One such system, free access stalls, allows the unique opportunity to investigate the preference of the animal for individual or group-housing. The objective of this study was to describe stall use by gestating gilts and parity 1 (P1) sows in a free access stall system. A gilt pen and a P1 pen, each with 8 stalls and 1 pig allocated per stall space (n = 16 total), were used for this study. Pigs were locked in the stalls 5 d before mixing (30 dg) to allow acclimation to the new pen environment and feeding schedule. On the day of mixing, females were moved out of their stalls at noon and continuous overhead video was captured for 72h post-mixing. To characterize stall use, videos were manually decoded sows time of entry into and exit from stalls. Time spent in stalls was analyzed using a Gaussian linear mixed model with fixed effects of parity, day, interaction between parity and day, and a random effect of individual animal. On average gilts spent significantly more time in stalls compared with P1 sows (20.9 h vs 16.4 h, P = 0.050). Overall, there was no change in stall utilization over the 3 d post-mixing but there was a significant interaction of parity and day (P = 0.008). On d 1, time spent in stalls did not differ between gilts and P1s (P = 0.743). However, on d 2 and 3 we observed significant differences between parities, with P1 sows spending less time in stalls than gilts (P = 0.032, P = 0.004, respectively). In conclusion, gestating females spent more time in stalls than in loafing areas. Ongoing research will explore how gestating females use their time in each space. The observed parity differences may suggest that P1 sows adapt more quickly to group housing systems compared with gilts.

Research on Intelligent Monitoring System of Intelligent Transmission Line Based on Computer Internet of Things Technology

The author proposes an intelligent early-warning system for preventing external damage of transmission lines. The system uses TI-TMS320DM6446 as the core, the peripheral circuit configuration and video monitoring; The moving objects captured by PTZ camera are detected and identified by using mixed Gaussian model, background difference and minimum tangent matrix, and alarm signals are provided. According to the particularity of the power supply of the equipment on the transmission line, the basic power management of the system is analyzed in detail. A reliable power management scheme is designed. The practical application proves that this method has good real-time performance and has obvious effect on preventing power cable damage by force.

Detection Method of Tourist Flow in Scenic Spots based on Kalman Filter Prediction

The tourism industry has developed rapidly, but it is always limited by the environmental carrying capacity and cannot receive too many tourists at the same time. Therefore, it is very necessary to limit the number of tourists visiting at the same time based on traffic detection. To this end, the tourist scenic spot (TSS) traffic statistics system was designed. The system performed graying, binarization, image denoising, and morphological processing on the image. The pre-processed image used the background difference method based on mixed Gaussian background modeling to detect moving objects. The improved Hough transform circle detection method was used to identify the head target, and the Kalman filter (KF) was used to complete the target tracking. KF could predict the target trajectory accurately, and the improved Hough transform circle detection method could recognize the head under occlusion. The maximum missed detection rate of the statistical system was 3.2\%, the minimum is 0, and the overall detection accuracy was the highest. The error rate of inbound passenger flow was 4.10\%, and the error rate of outbound passenger flow was 3.0\%. Using this system can control the tourist flow (TF) in the scenic spot and avoid safety accidents due to excessive passenger flow. And it is conducive to the sustainable development of the scenic spot.

From earthquakes to island area: multi‐scale effects upon local diversity

Tropical forests occupy small coral atolls to the vast Amazon basin. They occur across bioregions with different geological and climatic history. Differences in area and bioregional history shape species immigration, extinction and diversification. How this effects local diversity is unclear. The Indonesian archipelago hosts thousands of tree species whose coexistence should depend upon these factors. Using a novel dataset of 215 Indonesian forest plots, across fifteen islands ranging in area from 120 to 785 000 km2, we apply Gaussian mixed effects models to examine the simultaneous effects of environment, earthquake proximity, island area and bioregion upon tree diversity for trees ≥ 10 cm diameter at breast height. We find that tree diversity declines with precipitation seasonality and increases with island area. Accounting for the effects of environment and island area we show that the westernmost bioregion Sunda has greater local diversity than Wallacea, which in turn has greater local diversity than easternmost Sahul. However, when the model includes geological activity (here proximity to major earthquakes), bioregion differences are reduced. Overall, results indicate that multi‐scale, current and historic effects dictate tree diversity. These multi‐scale drivers should not be ignored when studying biodiversity gradients and their impacts upon ecosystem function.

Data Compression and Damage Evaluation of Underground Pipeline With Musicalized Sonar GMM

The underground cement pipeline network is critical in urban infrastructure, such as sewage drainage, fluid transportation, etc. Pipelines are deeply buried, so the damage is difficult to be detected. It is of great significant to monitor the crack of pipeline in real time. In this research, a data compression and damage evaluation method of underground pipeline using musicalized sonar GMM (Gaussian Mixed Model) was proposed. First, the sonar echo signal was sampled and discretized into four musicalized indicators by musicalized model. Then, the distribution of multi parameters were analyzed by GMM. The Frechet similarity was constructed to compare the difference between each working condition and baseline condition, so that the damage state could be identified. The engineering experiment shown that this method could greatly compress the monitoring data and evaluate four crack levels.

Ship collision avoidance decision-making research in coastal waters considering uncertainty of target ships

Ship collision avoidance has always been a concern and it is crucial for achieving safe navigation of ships at sea. There are many studies on ship collision avoidance in open water, but less attention on coastal waters considering the uncertainty of target ships due to the complexity of the environment and traffic flow. In this paper, collision avoidance decision-making research in coastal waters considering the uncertainty of target ships was proposed. Firstly, accurate ship trajectories are obtained by preprocessing the raw Automatic Identification System (AIS) data. Subsequently, the processed trajectories are clustered using the Ordering Points to Identify the Clustering Structure (OPTICS) algorithm and Hausdorff distance, acquiring a dataset for trajectory prediction of target ships. Then, a mixed Gaussian model is utilized to calculate the prior probability distribution of the prediction model, thus establishing a trajectory prediction model that considers the uncertainty of the target ship. Finally, ship maneuverability is simulated using the Mathematical Model Group (MMG) and Proportion Integration Differentiation (PID) models, and a collision avoidance decision-making model for ships is constructed. The proposed algorithm has been tested and verified in a case study. The results show that the approach effectively predicts the trajectory of the target ship and facilitates informed collision avoidance decision-making.

Research on lane-changing decision and control of autonomous vehicles based on game theory

Designing a secure and reliable decision-planning model for vehicle lane changing is of utmost practical significance because it is one of the most frequent driving behaviors and has a substantial impact on the safety of drivers’ lives and property. First, a Gaussian mixed Hidden Markov model (GMHMM) is trained for lane change intention recognition (LCIR), and the results reveal that the model has a great performance. This will simplify the game process and provide drivers and passengers with warnings. Second, the safety, efficiency, and comfort payoffs of vehicle lane changes are taken into account when building the game model. When building the safety payoff function, temporal collision risk and spatial collision risk of vehicles are two of them that are carefully taken into account. After that, the vehicle’s trajectory tracking control is decoupled into lateral LQR + feedforward control and longitudinal dual proportional integral derivative (PID) control based on the Frenet coordinate system. Finally, a vehicle lane change scenario is built for simulation analysis, and the effects of driving comfort factor and driving efficiency factor on lane change results are considered. The results show that the proposed game theory lane change model ensures lane change safety while satisfying human drivers’ requirements for lane change efficiency and comfort.

Comparison of brain functions between healthy participants and methamphetamine users with various addiction histories: Data analysis based on EEG and fNIRS

The electroencephalogram (EEG) rhythm and functional near-infrared spectroscopy (fNIRS) activation levels have not been compared between a healthy control group (HCG) and methamphetamine user group (MUG) with different addiction histories. This study used 64-electrode EEG and fNIRS to conduct an experiment that analyzed the resting and craving states. The EEG and fNIRS data of 56 participants were collected, including 14 healthy participants, 14 methamphetamine users with an addiction history of 0.5–5 years, 14 users with an addiction history of 5–10 years, and 14 users with an addiction history of 10–15 years. Isolated effective coherence (iCoh) within the brain network was used to process the EEG data. Statistical analysis was performed to compare differences in iCoh among the delta, theta, alpha, beta, and gamma bands and explore oxyhemoglobin activation levels in the ventrolateral prefrontal cortex, dorsolateral prefrontal cortex, orbitofrontal cortex, and frontopolar prefrontal cortex (FPC) of the control group. Finally, the Kmeans, Gaussian mixed model (GMM), linear discriminant analysis (LDA), support vector machine (SVM), Bayes, and convolutional neural networks (CNN) algorithms were used to classify methamphetamine users based on drug and neutral images. A 3-class accuracy was achieved. Changes in EEG and fNIRS activation levels of HCG and MUG with varied addiction histories were demonstrated.

Global vision object detection using an improved Gaussian Mixture model based on contour.

Object detection plays an important role in the field of computer vision. The purpose of object detection is to identify the objects of interest in the image and determine their categories and positions. Object detection has many important applications in various fields. This article addresses the problems of unclear foreground contour in moving object detection and excessive noise points in the global vision, proposing an improved Gaussian mixture model for feature fusion. First, the RGB image was converted into the HSV space, and a mixed Gaussian background model was established. Next, the object area was obtained through background subtraction, residual interference in the foreground was removed using the median filtering method, and morphological processing was performed. Then, an improved Canny algorithm using an automatic threshold from the Otsu method was used to extract the overall object contour. Finally, feature fusion of edge contours and the foreground area was performed to obtain the final object contour. The experimental results show that this method improves the accuracy of the object contour and reduces noise in the object.

An objective evaluation method of vocal singing effect based on artificial intelligence technology

Abstract The continuous progress of artificial intelligence technology has shown great potential for application in several fields, especially music. The research direction of Objective Evaluation of Vocal Singing Effectiveness uses advanced technologies to analyze and assess a singer’s performance across multiple dimensions, including pitch, rhythm, and timbre, and is highly valuable. Building an accurate and fair evaluation system faces many challenges, including how to accurately capture and analyze the subtle changes in the voice and synthesize the effects of different musical elements on the quality of the performance. This requires researchers to explore music theory, sound analysis techniques, and artificial intelligence algorithms, and develop a new methodology that can comprehensively evaluate the effectiveness of vocal singing. This paper constructs a complete set of vocal singing evaluation models by analyzing acoustic feature extraction, Hidden Markov Model, and Generalized Regression Radial Basis Function Network in detail. The study adopts a logarithmic Mel spectrum for acoustic feature extraction to effectively capture the essential attributes of the singing voice. Hidden Markov models and mixed Gaussian models are used to model the sound signal, improving phoneme recognition accuracy. Accurate singing effect was evaluated using a generalized regression radial basis function network. In this article, the accuracy of this evaluation method in terms of pitch, rhythm, and timbre reached 95%, 93%, and 89%, respectively, demonstrating high evaluation consistency and reliability. The research method provides a new objective evaluation tool for vocal singing effects, which is valuable for vocal teaching and self-practice.

Adaptive region prediction of gravity aided navigation system based on SVM multi-classification and mixed Gaussian clustering model

Adaptive region prediction of gravity aided navigation system based on SVM multi-classification and mixed Gaussian clustering model

Thermal runaway detection method of SiO<SUB align="right">2 aerogel battery pack based on mixed Gaussian hidden Markov model

Thermal runaway detection method of SiO<SUB align="right">2 aerogel battery pack based on mixed Gaussian hidden Markov model