Recurrent Neural Network and Bionic Wavelet Transform for speech enhancement

Talbi, Mourad; Lotfi, Shahriar; Abid, S.; Chérif, Adnane

doi:10.1504/ijsise.2010.035002

Cited by 4 publications

(5 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As shown in this figure, we first apply the bionic wavelet transform (BWT) to the noisy ECG signal. The application of the BWT is performed by using 30 scales (Mourad et al 2010) instead of 22 scales (Yao & Zhang 2001;Sayadi & Shamsollahi 2006).…”

Section: New Proposed De-noising Techniquementioning

confidence: 99%

Electrocardiogram de-noising based on forward wavelet transform translation invariant application in bionic wavelet domain

Talbi

2014

Sadhana

View full text Add to dashboard Cite

In this paper, we propose a new technique of Electrocardiogram (ECG) signal de-noising based on thresholding of the coefficients obtained from the application of the Forward Wavelet Transform Translation Invariant (FWT_TI) to each Bionic Wavelet coefficient. The De-noise De-noised ECG is obtained from the application of the inverse of BWT (BW T −1) to the de-noise de-noised bionic wavelet coefficients. For evaluating this new proposed de-noising technique, we have compared it to a thresholding technique in the FWT_TI domain. Preliminary tests of the application of the two de-noising techniques were constructed on a number of ECG signals taken from MIT-BIH database. The obtained results from Signal to Noise Ratio (SNR) and Mean Square Error (MSE) computations showed that our proposed de-noising technique outperforms the second technique. We have also compared the proposed technique to the thresholding technique in the bionic wavelet domain and this comparison was performed by SNR improvement computing. The obtained results from this evaluation showed that the proposed technique also outperforms the de-noising technique based on bionic wavelet coefficients thresholding.

show abstract

Section: New Proposed De-noising Techniquementioning

confidence: 99%

Electrocardiogram de-noising based on forward wavelet transform translation invariant application in bionic wavelet domain

Talbi

2014

Sadhana

View full text Add to dashboard Cite

show abstract

“…Hence, the adaptive nature of the BWT is captured by a time-varying factor T. This factor represents the scaling of the cochlear filter bank quality at each scale over time [12][13][14][15][16]. For the human auditory system, Yao and Zhang [12,13] have taken   = 15165.4 .…”

Section: Bionic Wavelet Transformmentioning

confidence: 99%

“…For the human auditory system, Yao and Zhang [12,13] have taken   = 15165.4 . The discrimination of the scale variable  is accomplished using a pre-defined logarithmic spacing across the desired frequency rang so that the center frequency at each scale is expressed as follows [14][15][16]:…”

Section: Bionic Wavelet Transformmentioning

confidence: 99%

“…The 21 scales correspond to center frequencies logarithmically spaced from 166.4 Hz to 3369.7 Hz. For each time and scale, the adapting function (, ) is calculated using the following equation [14,15,16]:…”

Section: Bionic Wavelet Transformmentioning

confidence: 99%

“…In this paper, we employ the same values as in the reference [13][14][15][16]:    = 0.87,    = 45,    = 0.8 and   = 0.0005 . Finally, the computation step ∆ is chosen to be equal to 1   ⁄ , where   represents the sampling frequency.…”

Section: Bionic Wavelet Transformmentioning

confidence: 99%

See 2 more Smart Citations

Wiener Filtering Application in the Bionic Wavelet Domain for Speech Enhancement

Talbi¹,

Lotfi²,

Mouhamed³

et al. 2012

IJACT

View full text Add to dashboard Cite

In this paper, a new speech enhancement method is proposed and is based on the application of the Wiener filtering in the Bionic Wavelet Transform (BWT) domain. The BWT provides a time-frequency selectivity and better energy concentration property. The proposed method was compared to a number of speech enhancement methods including a BWT based speech enhancement method, Wiener filtering ,and MMSE-STSA. Results are evaluated objectively using signal-to-noise ratio (SNR), segmental signal-to-noise ratio (segSNR), Itakura-Saito distance (ISd) and perceptual evaluation of speech quality (PESQ) with TIMIT sentences corrupted by various noise types and SNRs. The obtained results show that the proposed method gives good results when compared to the other techniques.

show abstract

Speech enhancement based on stationary bionic wavelet transform and maximum a posterior estimator of magnitude-squared spectrum

Talbi

2016

Int J Speech Technol

View full text Add to dashboard Cite

Recurrent Neural Network and Bionic Wavelet Transform for speech enhancement

Cited by 4 publications

References 0 publications

Electrocardiogram de-noising based on forward wavelet transform translation invariant application in bionic wavelet domain

Electrocardiogram de-noising based on forward wavelet transform translation invariant application in bionic wavelet domain

Wiener Filtering Application in the Bionic Wavelet Domain for Speech Enhancement

Speech enhancement based on stationary bionic wavelet transform and maximum a posterior estimator of magnitude-squared spectrum

Contact Info

Product

Resources

About