Related Experiment Video
Updated: Mar 15, 2026
09:09
Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody
Published on: September 27, 2024
957
High-order hidden Markov model for piecewise linear processes and applications to speech recognition
1Department of Electrical Engineering, Da-Yeh University, 168 University Road, Dacun, Changhua, Taiwan.
The Journal of the Acoustical Society of America
|September 3, 2016
Summary
A novel high-order hidden Markov model (HMM) improves piecewise linear process approximation. This advanced HMM reduces speech recognition error rates for noisy Mandarin digits compared to standard HMMs.
Area of Science:
- Machine Learning
- Signal Processing
- Speech Recognition
Background:
- Hidden Markov Models (HMMs) are standard for sequential data.
- Standard HMMs struggle with real-world processes due to output conditional independence, limiting them to piecewise constant sequences.
- Accurate modeling of complex real processes requires more flexible sequence generation.
Purpose of the Study:
- To propose a high-order hidden Markov model (HMM) capable of modeling piecewise linear processes.
- To develop a parameter estimation method for the new HMM using the expectation-maximization algorithm.
- To evaluate the effectiveness of the proposed HMM in noisy speech recognition tasks.
Main Methods:
- Development of a high-order hidden Markov model (HMM) architecture.
- Implementation of a parameter estimation technique based on the expectation-maximization (EM) algorithm.
- Experimental validation using speech recognition of noisy Mandarin digits.
Main Results:
- The proposed high-order HMM effectively models piecewise linear processes.
- The expectation-maximization algorithm successfully estimated model parameters.
- Experimental results demonstrated a reduced recognition error rate compared to a baseline HMM.
Conclusions:
- The high-order HMM offers a superior approximation for piecewise linear processes compared to standard HMMs.
- The developed parameter estimation method is effective for the proposed model.
- The advanced HMM shows significant potential for improving speech recognition accuracy in noisy environments.
Related Concept Videos
01:29Integration by Parts: Problem Solving
105
Smart speakers process voice commands by modeling audio inputs as piecewise functions and analyzing them through integration against trigonometric functions, such as cosine. This mathematical approach is fundamental in signal processing, where complex sound waves are decomposed into simpler frequency components.Consider a definite integral involving a piecewise function multiplied by a cosine function. Because the function is defined differently over separate intervals, the integral is split...
105
01:29Application of Linearization and Approximation
129
A drone flying through complex terrain often relies on more than one sensing method to estimate small changes in altitude. Along with direct measurements, air pressure provides a useful indirect indicator of vertical movement. Atmospheric pressure decreases as altitude increases, and this relationship is commonly described using an exponential model. Although accurate, converting pressure measurements into altitude values requires calculations that are too complex to perform repeatedly during...
129
01:28Types of Functions III
319
Logarithmic and piecewise functions play central roles in mathematical modeling, particularly when capturing nonlinear or segmented behaviors in real-world phenomena. Although these functions differ fundamentally in structure and application, both serve to represent complex relationships in simplified mathematical terms.A logarithmic function is defined as the inverse of an exponential function, expressed as These functions grow quickly for small values of x but slow down as x increases,...
319
01:26Linearization and Approximation
143
Linearization is a mathematical technique used to approximate complex, nonlinear functions with simpler linear models in the vicinity of a chosen reference point. The method is based on the idea that, although a function may be difficult to evaluate exactly, its behavior near a specific input value can often be closely approximated by the tangent line at that point. This approach is particularly useful when small deviations from a known value are involved.Consider the square root function, for...
143
01:27State Space Representation
658
The frequency-domain technique, commonly used in analyzing and designing feedback control systems, is effective for linear, time-invariant systems. However, it falls short when dealing with nonlinear, time-varying, and multiple-input multiple-output systems. The time-domain or state-space approach addresses these limitations by utilizing state variables to construct simultaneous, first-order differential equations, known as state equations, for an nth-order system.
Consider an RLC circuit, a...
Consider an RLC circuit, a...
658
01:26Linear Approximation in Frequency Domain
418
Linear systems are characterized by two main properties: superposition and homogeneity. Superposition allows the response to multiple inputs to be the sum of the responses to each individual input. Homogeneity ensures that scaling an input by a scalar results in the response being scaled by the same scalar.
In contrast, nonlinear systems do not inherently possess these properties. However, for small deviations around an operating point, a nonlinear system can often be approximated as linear....
In contrast, nonlinear systems do not inherently possess these properties. However, for small deviations around an operating point, a nonlinear system can often be approximated as linear....
418