Explore MATLAB source code curated for "语音识别" with clean implementations, documentation, and examples.
This document contains various algorithms used in speech recognition systems, including frame-based processing algorithms, MFCC (Mel-Frequency Cepstral Coefficients) computation, and cepstral analysis algorithms, all implemented using MATLAB. The package also includes an audio sample file for testing and demonstration purposes.
MFCC parameter extraction is crucial in speech recognition systems, offering both efficient storage and compatibility with human auditory perception. This MATLAB simulation code provides complete and high-performance feature extraction functionality, implementing the standard MFCC algorithm with optimized signal processing techniques.
MATLAB code for speech recognition including test audio samples, featuring implementation of both HMM (Hidden Markov Model) and DTW (Dynamic Time Warping) methods with detailed algorithm explanations and practical applications.
A comprehensive and practical implementation of Gaussian Mixture Model (GMM) for speech recognition, featuring detailed code structure and algorithm explanations
Voice Activity Detection (VAD) serves as a fundamental technology in speech recognition with extensive applications across both commercial and civil domains. Accurate endpoint detection becomes particularly challenging in low signal-to-noise ratio (SNR) environments, especially during silent segments or transitional periods before and after phonation.
MATLAB source code for a complete HMM-based speech recognition system featuring feature extraction, model training, and pattern recognition algorithms
Speech Recognition with Energy Computation, Zero-Crossing Rate Analysis, and Onset Detection Techniques for Musical Note Segmentation
Speech recognition system implementation with GUI interface for digit recognition (0-9), featuring signal processing algorithms and machine learning techniques
This repository contains the original author's implementation code for Two-Dimensional PCA (2DPCA) - a novel approach for appearance-based face representation and recognition published in IEEE PAMI. The paper has received 681 citations on Google Scholar. Note: While this algorithm is a classic in pattern recognition (primarily for facial recognition), the "speech recognition" categorization is inaccurate as 2DPCA is not commonly applied to audio processing. The code provides practical implementation of 2DPCA's matrix-based dimensionality reduction technique.
This example implements a speech recognition system in MATLAB using traditional MFCC feature extraction and popular SVM classifier, serving as an excellent baseline for comparative studies. The implementation includes feature extraction pipeline and classification algorithms suitable for benchmarking purposes.