GMM模型 Resources

Showing items tagged with "GMM模型"

This paper presents an effective implementation of Gaussian Mixture Models (GMM), a classic speaker recognition algorithm, using the Expectation-Maximization (EM) algorithm. The study primarily simulates GMM's noise robustness performance under various acoustic environments, yielding valuable insights for practical applications. Key implementation aspects include parameter initialization strategies and convergence criteria for the EM iteration process.

MATLAB 291 views Tagged

A speech emotion recognition system based on the Gaussian Mixture Model (GMM) framework, where GMM serves as a mathematical model for fitting data distributions. Discrepancies between observed data patterns and model outputs are expected since EM algorithm estimation of GMM parameters typically assumes incomplete data - meaning the algorithm computationally "completes" hidden or missing data components during parameter optimization. The system implementation involves feature extraction from speech signals, GMM parameter initialization, iterative EM updates for mean vectors, covariance matrices, and mixture weights, followed by maximum likelihood classification for emotion categorization.

MATLAB 279 views Tagged

MATLAB implementation for GMM speaker recognition model training, requiring integration with Voicebox toolbox for MFCC feature extraction and k-means clustering initialization.

MATLAB 309 views Tagged