A comparison of procedures for the calculation of forensic likelihood ratios from acoustic–phonetic data: Multivariate kernel density (MVKD) versus Gaussian mixture model–universal background model (GMM–UBM)

2011 ◽  
Vol 53 (2) ◽  
pp. 242-256 ◽  
Author(s):  
Geoffrey Stewart Morrison
Keyword(s):  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Kernel Density ◽  
Gaussian Mixture ◽  
Background Model ◽  
Likelihood Ratios ◽  
Universal Background Model

scholarly journals GMM-UBM Based Modeling for Language Identification using New Feature Vectors

2020 ◽  
Vol 9 (4) ◽  
pp. 3034-3039
Keyword(s):  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Gaussian Mixture ◽  
Language Identification ◽  
Training Data ◽  
Background Model ◽  
Indian Languages ◽  
Feature Vectors ◽  
Universal Background Model ◽  
New Feature

The most of the existing LID systems based on the Gaussian Mixture model. The main requirement of the GMM based LID system is it require large amount of speech data to train the GMM model. Most of the Indian languages have the similarity because they are derived from Devanagari. Even though common phonemes exists in phoneme sets across the Indian languages, each language contain its unique phonotactic constraints imposed by the language. Any modeling technique capable of capturing all these slight variations imposed by the language is one of the important language identification cue. To model the GMM based LID system which captures above variations it require large number of mixture components.To model the large number of mixture components using Gaussian Mixture Model (GMM), the technique requires a large number of training data for each language class, which is very difficult to get for Indian languages. The main objective of GMM-UBM based LID system is it require less amount of training data to train(model) the system. In this paper, the importance of GMM-UBM modeling for language identification (LID) task for Indian languages are explored using new set of feature vectors. In GMM-UBM LID system based on the new feature vectors, the phonotactic variations imparted by different Indian languages are modeled using Gaussian Mixture model and Universal Background Model (GMM-UBM) technique. In this type of modeling, some amount of data from each class of language is pooled to create a universal background model. From this UBM model each model class is adapted. In this study, it is found that the performance of new feature vectors GMM-UBM based LID system is superior when compared to conventional new feature vectors based GMM LID system.

scholarly journals A Review on Text-Independent Speaker Verification Techniques in Realistic World

2016 ◽  
Vol 9 (1) ◽  
pp. 36-40
Author(s):  
Renu Singh ◽  
Arvind Singh ◽  
Utpal Bhattacharjee
Keyword(s):  
Support Vector Machine ◽  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Speaker Verification ◽  
Gaussian Mixture ◽  
Background Model ◽  
Support Vector ◽  
Universal Background Model ◽  
Verification Techniques ◽  
Text Independent Speaker Verification

This paper presents a reviewof various speaker verification approaches in realistic world, and explore a combinational approach between Gaussian Mixture Model (GMM) and Support Vector Machine (SVM) as well as Gaussian Mixture Model (GMM) and Universal Background Model (UBM).

Algorithm of Moving Vehicle Detection by Improved Gaussian Mixture Model Based on ROI

2015 ◽  
Vol 734 ◽  
pp. 463-467 ◽  
Author(s):  
Pan Pan Zhang ◽  
Chun Yang Mu ◽  
Xing Ma ◽  
Fu Lu Xu
Keyword(s):  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Region Of Interest ◽  
Gaussian Mixture ◽  
Background Model ◽  
Morphological Filter ◽  
Improved Method ◽  
Moving Vehicle ◽  
Model Based ◽  
Moving Vehicle Detection

Detection of moving object is a hot topic in computer vision. Traditionally, it is detected for every pixel in whole image by Gaussian mixture background model, which may waste more time and space. In order to improving the computational efficiency, an advanced Gaussian mixture model based on Region of Interest was proposed. Firstly, the solution finds out the most probably region where the target may turn up. And then Gaussian mixture background model is built in this area. Finally, morphological filter algorithm is used for improving integrity of the detected targets. Results show that the improved method could have a more perfect detection but no more time increasing than typical method.

Research on Background Modeling by Gaussian Mixture Model without Parameter Adjustment

2013 ◽  
Vol 380-384 ◽  
pp. 1394-1397
Author(s):  
Hong Hai Liu ◽  
Xiang Hua Hou
Keyword(s):  
Density Estimation ◽  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Mean Shift ◽  
Gaussian Mixture ◽  
Background Modeling ◽  
Background Model ◽  
Probability Density Estimation ◽  
Mean Shift Algorithm ◽  
Statistical Background

When modeling background model by Gaussian mixture model, there exist the defects that parameters can not be updated adaptively. In this paper, we adopt mean-shift algorithm to overcome these defects. Firstly, this paper introduces the initialized parameters, such as variance, mean, and weights and others, when modeling and then the parameters are constantly adjusted in the subsequent calculations. Then the statistical background model based on probability density estimation is put forward and using mean-shift algorithm updates the parameters adaptively. At last, the algorithm of mixture Gaussian background modeling method based on mean-shift is implemented. The experimental results show that the algorithm can effectively update parameters adaptively and the obtained background model is better.

scholarly journals Evaluating the Effect of Intensity Standardisation on Longitudinal Whole Brain Atrophy Quantification in Brain Magnetic Resonance Imaging

2020 ◽  
Author(s):  
Emily Esmeralda Carvajal-Camelo ◽  
Jose Bernal ◽  
Arnau Oliver ◽  
Xavier Lladó ◽  
Maria Trujillo
Keyword(s):  
Magnetic Resonance Imaging ◽  
Magnetic Resonance ◽  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Brain Magnetic Resonance Imaging ◽  
Kernel Density ◽  
Gaussian Mixture ◽  
Resonance Imaging ◽  
Fuzzy C Means ◽  
Skull Stripping

Atrophy quantification is fundamental for understanding brain development and diagnosing and monitoring brain diseases. FSL-SIENA is a well-known fully-automated method that has been widely used in brain magnetic resonance imaging studies. However, intensity variations arising during image acquisition that may compromise evaluation, analysis and even diagnosis. In this work, we study whether intensity standardisation can improve longitudinal atrophy quantification. We considered seven methods comprising z-score, fuzzy c-means, Gaussian mixture model, kernel density, histogram matching, white stripe, and removal of artificial voxel effects by linear regression (RAVEL). We used a total of 330 scans from two publicly-available datasets, ADNI and OASIS. In scan-rescan assessments, that measures robustness to subtle imaging variations, intensity standardisation did not compromise the robustness of FSL-SIENA significantly (p>0.1). In power analysis assessments, that measures the ability to discern between two groups of subjects, three methods led to consistent improvements in both datasets with respect to the original: fuzzy c-means, Gaussian mixture model, and kernel density estimation. Reduction in sample size using these three methods ranged from 17% to 95%. The performance of the other four methods was affected by spatial normalisation, skull stripping errors, presence of periventricular white matter hyperintensities, or tissue proportion variations over time. Our work evinces the relevance of appropriate intensity standardisation in longitudinal cerebral atrophy assessments using FSL-SIENA.

Sign in / Sign up

Export Citation Format

Share Document