Enhanced Speech Based Jointly Statistical Probability Distribution Function for Voice Activity Detection

Li, Jie; You, Datao

doi:10.1049/cje.2017.01.001

Cited by 5 publications

(3 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the process of front-end processing, the Voice activity detection (VAD) [18] technology is used to remove the mute in the speech. And then the experiments operated on cepstral features which was extracted using a 32-ms Hamming window.…”

Section: Methodsmentioning

confidence: 99%

Partial Least Squares Based Total Variability Space Modeling for I‐Vector Speaker Verification

Chen

Han

2018

Chin. j. electron.

View full text Add to dashboard Cite

Section: Methodsmentioning

confidence: 99%

Partial Least Squares Based Total Variability Space Modeling for I‐Vector Speaker Verification

Chen

Han

2018

Chin. j. electron.

View full text Add to dashboard Cite

“…The methods often hypothesize the discrete Fourier transform (DFT) coefficients of noise is single Gaussian distribution for simplifying the calculation, although it is not the case in practice. This problem leads to the methods based on statistical model give a bad distinguishing ability in low SNR especially in non-stationary noisy conditions [44]. Besides, the complicated computation of the method increases the time cost and is not conducive to implementing on smartphone in real time.…”

Section: B Voice Activity Detection (Vad)mentioning

confidence: 99%

HCI on the Table: Robust Gesture Recognition Using Acoustic Sensing in Your Hand

et al. 2020

View full text Add to dashboard Cite

The paper proposes a new HCI mechanism for device-free gesture recognition on the table using acoustic signal, which can extend the gesture input and interactions beyond the tiny screen of mobile device and allow users to provide input without blocking screen view. Previous researches have either relied on additional devices (e.g., special wearable device and mouse) or required active acoustic signals which demand additional cost and be less prone to popularize, while we explore the device-free gesture recognition using passive acoustic signals. This technology is more challenging due to the lack of an effective approach to eliminate the inherent ambient noise disturbances and extract stable gesture features. We fuse both short time energy (STE) and zero-crossing rate (ZCR) to identify the effective signals from the original input, and leverage the Mel frequency cepstral coefficients (MFCC), cochlear filter cepstral coefficients (CFCC) to extract the stable features from different gestures. The unique features in support vector machine (SVM) classifier achieve a high gesture recognition accuracy from the noisy scenarios and mismatched conditions. Implementation on the Android system has realized real-time processing of the feature extraction and gesture recognition. Extensive evaluations show our algorithm has a better noise tolerant performance and the system could recognize seven common gestures (click, flip left/right, scroll up/down, zoom in/out) on smart devices with an accuracy of 93.2%.

show abstract

“…e Laplacian distribution is used to model the residual noise, since the residual noise in the enhanced speech satisfies the Laplacian distribution. Experimental results showed that his proposed method performs better than baseline methods, especially under low SNR and nonstationary noise conditions [6]. Xue et al proposed a vision-centric multisensor fusion framework for traffic environment perception methods for autonomous driving.…”

Section: Introductionmentioning

confidence: 99%

Multisensor Speech Enhancement Technology in Music Synthesizer Design

Peng

2022

Mobile Information Systems

View full text Add to dashboard Cite

Creating music through sound synthesis is the most representative electronic music creation method, and electronic music is actually the result of sound synthesis technology. Today, the field of electronic music encompasses multiple areas such as recording, mixing, composing, and producing. It also has some advantages over traditional music composition. Voice is the most effective and direct way of communication between people. And with the explosive development of speech recognition technology, the recognition rate of speech recognition systems in the near field environment has been greatly improved. However, in practical applications, there is often a large amount of ambient noise. If these environmental noises are strong, it will seriously affect the quality, accuracy, and speed of music synthesis. This greatly reduces not only the sound quality and clarity of speech but also the speed of speech recognition. To solve these problems, this paper proposes a multisensor speech enhancement technique and implements a multisensor speech enhancement system. It also proposes an enhancement method based on speaker speech and microphone speech. In this paper, the low-frequency harmonic components of the bone conduction signal are used to replace the frequency points disturbed by wind noise to reduce the influence of wind noise on speech quality and intelligibility. The experimental results show that the PESQ and MOS scores of the improved algorithm in this paper are 1.65 and 3.67, respectively. Compared with the existing methods, it has a great improvement. This can effectively improve the voice quality of the music synthesizer and reduce background noise.

show abstract

Enhanced Speech Based Jointly Statistical Probability Distribution Function for Voice Activity Detection

Cited by 5 publications

References 13 publications

Partial Least Squares Based Total Variability Space Modeling for I‐Vector Speaker Verification

Partial Least Squares Based Total Variability Space Modeling for I‐Vector Speaker Verification

HCI on the Table: Robust Gesture Recognition Using Acoustic Sensing in Your Hand

Multisensor Speech Enhancement Technology in Music Synthesizer Design

Contact Info

Product

Resources

About