Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Probability Histograms01:17

Probability Histograms

13.3K
A probability histogram is a visual representation of a probability distribution. Similar a typical histogram, the probability histogram consists of contiguous (adjoining) boxes. It has both a horizontal axis and a vertical axis. The horizontal axis is labeled with what the data represents. The vertical axis is labeled with probability. Each rectangular bar in the histogram is 1 unit wide, which suggests that the area under each bar equals the probability, P(x), where x is 1, 2, 3, and so on.
13.3K
Histogram01:05

Histogram

18.1K
The histogram is a graphical representation in the x-y form of data distribution in a data set. The horizontal x-axis is labeled with what the data represents (for instance, distance from your home to school). The vertical y-axis is labeled either frequency or relative frequency (or percent frequency or probability).
A histogram graph consists of contiguous (adjoining) boxes. The heights of the bars correspond to frequency values. The graph will have the same shape with respective labels. The...
18.1K
One-Way ANOVA: Equal Sample Sizes01:15

One-Way ANOVA: Equal Sample Sizes

4.2K
One-Way ANOVA can be performed on three or more samples with equal or unequal sample sizes. When one-way ANOVA is performed on two datasets with samples of equal sizes, it can be easily observed that the computed F statistic is highly sensitive to the sample mean.
Different sample means can result in different values for the variance estimate: variance between samples. This is because the variance between samples is calculated as the product of the sample size and the variance between the...
4.2K
Relative Frequency Histogram01:14

Relative Frequency Histogram

6.5K
The relative frequency depicts the proportion of data points that have each value. The frequency tells the number of data points that have each value. Like the histogram, a relative frequency histogram also has the same shape with a horizontal scale (the x-axis), but the vertical scale (the y-axis) is marked with relative frequencies (percentages of the whole) instead of actual frequencies. A relative frequency histogram is a graphical representation of a frequency distribution where the...
6.5K
Electric Field of Two Equal and Opposite Charges01:30

Electric Field of Two Equal and Opposite Charges

7.1K
Atoms generally contain the same number of positively and negatively charged particles, protons, and electrons. Hence, they are electrically neutral. However, the centers of the positive and negative charges do not always coincide. In such a scenario, the electric field of an atom may not be zero.
A separation of the positive and negative charges can lead to a weak, remnant effect of the positive and negative charges. The expectation is that the more the distance between the positive and...
7.1K
What are Estimates?01:06

What are Estimates?

8.9K
It isn't easy to measure a parameter such as the mean height or the mean weight of a population. So, we draw samples from the population and calculate the mean height or mean weight of the individuals in the sample. This sample data acts as a representative measure of the population parameter. These sample statistics are known as estimates. 
The estimate for the mean of a sample is denoted by ͞x, whereas the mean of the population is designated as μ. Further, parameters such...
8.9K

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Acoustic biomarkers for schizophrenia spectrum disorders and their associations with symptoms and cognitive functioning.

Progress in neuro-psychopharmacology & biological psychiatry·2025
Same author

Regularized Speaker Adaptation of KL-HMM for Dysarthric Speech Recognition.

IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society·2017
Same author

Minimum classification error-based weighted support vector machine kernels for speaker verification.

The Journal of the Acoustical Society of America·2013
See all related articles

Related Experiment Video

Updated: Feb 13, 2026

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception
05:48

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception

Published on: August 9, 2024

2.0K

Histogram equalization with Bayesian estimation for noise robust speech recognition.

Youngjoo Suh1, Hoirin Kim1

  • 1School of Electrical Engineering, Korea Advanced Institute of Science and Technology, 291 Daehak-ro, Yuseong-gu, Daejeon, 34141, Republic of Korea.

The Journal of the Acoustical Society of America
|March 3, 2018
PubMed
Summary

This study introduces a Bayesian estimation method for histogram equalization to improve automatic speech recognition (ASR) systems. The enhanced technique boosts performance in deep neural network-hidden Markov model (DNN-HMM) systems, especially when data is limited.

More Related Videos

A Robust Single-Particle Cryo-Electron Microscopy cryo-EM Processing Workflow with cryoSPARC, RELION, and Scipion
13:43

A Robust Single-Particle Cryo-Electron Microscopy cryo-EM Processing Workflow with cryoSPARC, RELION, and Scipion

Published on: January 31, 2022

15.3K
Ultrasound Images of the Tongue: A Tutorial for Assessment and Remediation of Speech Sound Errors
08:32

Ultrasound Images of the Tongue: A Tutorial for Assessment and Remediation of Speech Sound Errors

Published on: January 3, 2017

23.3K

Related Experiment Videos

Last Updated: Feb 13, 2026

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception
05:48

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception

Published on: August 9, 2024

2.0K
A Robust Single-Particle Cryo-Electron Microscopy cryo-EM Processing Workflow with cryoSPARC, RELION, and Scipion
13:43

A Robust Single-Particle Cryo-Electron Microscopy cryo-EM Processing Workflow with cryoSPARC, RELION, and Scipion

Published on: January 31, 2022

15.3K
Ultrasound Images of the Tongue: A Tutorial for Assessment and Remediation of Speech Sound Errors
08:32

Ultrasound Images of the Tongue: A Tutorial for Assessment and Remediation of Speech Sound Errors

Published on: January 3, 2017

23.3K

Area of Science:

  • Speech Recognition
  • Machine Learning
  • Signal Processing

Background:

  • Histogram equalization is an efficient feature normalization technique for noise-robust automatic speech recognition (ASR).
  • Conventional methods degrade performance under specific test conditions or due to overfitting with insufficient data.
  • Class-based histogram equalization improved noise robustness but still faced overfitting issues.

Purpose of the Study:

  • To address the limitations of existing histogram equalization methods in ASR.
  • To improve the performance of speech recognition systems, particularly those using deep neural network-hidden Markov models (DNN-HMMs).
  • To enhance feature normalization techniques for better robustness and accuracy in diverse acoustic environments.

Main Methods:

  • The proposed histogram equalization technique utilizes Bayesian estimation for cumulative distribution function estimation.
  • The method was evaluated on speech recognition systems, including Gaussian mixture model-hidden Markov models (GMM-HMMs) and deep neural network-hidden Markov models (DNN-HMMs).
  • Fusion of proposed features with mel-frequency cepstral coefficients (MFCCs) was explored.

Main Results:

  • The Bayesian estimation approach demonstrated substantial performance gains in GMM-HMM based systems on the Aurora-4 task.
  • Meaningful performance improvements were observed in DNN-HMM systems compared to conventional maximum likelihood estimation methods.
  • Fusion with MFCCs provided additional gains in DNN-HMM systems, mitigating degradation in clean test conditions.

Conclusions:

  • The proposed Bayesian estimation-based histogram equalization is effective for enhancing ASR performance, particularly in DNN-HMM systems.
  • This technique offers a solution to overfitting problems and improves robustness in challenging acoustic environments.
  • Feature fusion further boosts performance, highlighting the adaptability of the proposed method to current ASR architectures.