In this paper, a single-channel speech enhancement algorithm is proposed by using guided spectrogram filtering based on masking properties of human auditory system when considering a speech spectrogram as an image. Guided filtering is capable of sharpening details and estimating unwanted textures or background noise from the noisy speech spectrogram. If we consider the noisy spectrogram as a degraded image, we can estimate the spectrogram of the clean speech signal using guided filtering after subtracting noise components. Combined with masking properties of human auditory system, the proposed algorithm adaptively adjusts and reduces the residual noise of the enhanced speech spectrogram according to the corresponding masking threshold. Because the filtering output is a local linear transform of the guidance spectrogram, the local mask window slides can be efficiently implemented via box filter with O(N) computational complexity. Experimental results show that the proposed algorithm can effectively suppress noise in different noisy environments and thus can greatly improve speech quality and speech intelligibility.
BenestyJ., Introduction. In Fundamentals of speech enhancement; Springer: Berlin, Germany, 2018; pp. 1–3.
2.
ZhangX., Principle and application of speech enhancement. In Digital audio processing and MATLAB simulation, 2th ed.; PHEI: Bejing, China, 2016; pp. 199–218.
3.
LoizouP.C., Noise-estimation algorithms. In Speech enhancement: theory and practice, 2th ed.; CRC Press: Boca Raton, USA, 2013; pp. 377–433.
4.
CohenI., BenestyJ.J. and GannotS., Simultaneous detection and estimation approach for speech enhancement and interference suppression. In Fundamentals of speech enhancement; Springer: Berlin, Germany, 2010; pp. 127–149.
5.
ErkelensJ.S., HendriksR.C., HeusdensR. and JensenJ., Minimum mean-square error estimation of discrete Fourier coefficients with generalized Gamma priors, IEEE Trans Audio, Speech, Language Process15 (2007), 1741–1752.
6.
SayoudA., DjendiM. and GuessoumA., A new speech enhancement adaptive algorithm based on fullband–subband MSE switching, J Speech Technol22 (2019), 865–884.
7.
LiN. and LoizouP.C., Factors influencing intelligibility of ideal binary-masked speech: Implications for noise reduction, J Acoust Soc Am123 (2008), 1673–1682.
8.
LoizouP.C. and KimG., Reasons why current speech-enhancement algorithms do not improve speech intelligibility and suggested solutions, IEEE Trans Audio, Speech, Language Process19 (2011), 47–56.
9.
LiJ., YangL., ZhangJ., YanY., HuY. and LoizouP.C., Comparative intelligibility investigation of single-channel noise-reduction algorithms for Chinese, Japanese, and English, J Acoust Soc Am129 (2011), 3291–3301.
10.
WangJ., LiuH., ZhengC. and LiX., Spectral subtraction based on two-stage spectral estimation and modified cepstrum thresholding, Acoust74 (2013), 450–458.
11.
LauferY., Laufer-GoldshteinB. and GannotS., ML estimation and CRBs for reverberation, speech, and noise PSDs in rank-deficient noise field, IEEE Trans Audio, Speech, Language Process28 (2020), 619–634.
12.
ZhangQ., NicolsonA.M., WangM., et al., Deep MMSE: A deep learning approach to MMSE-based noise power spectral density estimation, IEEE Trans Audio, Speech, Language Process 2020, pp. 1–1.
13.
GerkmannT. and HendriksR.C., Unbiased MMSE-based noise power estimation with low complexity and low tracking delay, IEEE Trans Audio, Speech, Language Process20 (2011), 1383–1393.
14.
XuY., DuJ., DaiL.R. and LeeC.H., An experimental study on speech enhancement based on deep neural networks, IEEE Signal Processing Lett21 (2013), 65–68.
15.
XuY., DuJ., DaiL.R. and LeeC.H., A regression approach to speech enhancement based on deep neural networks, IEEE Trans Audio, Speech, Language Process23 (2015), 7–19.
16.
KolbkM., TanZ.H. and JensenJ., Speech intelligibility potential of general and specialized deep neural network based speech enhancement systems, IEEE Trans Audio, Speech, Language Process25 (2017), 153–167.
17.
PandeyA. and WangD.L., A New Framework for CNN-Based Speech Enhancement in the Time Domain, IEEE Trans Audio, Speech, Language Process27 (2019), 1179–1188.
18.
HeK., SunJ. and TangX., Guided image filtering, IEEE Trans Pattern Anal Machine Intell35 (2013), 1397–1409.
19.
HaoS., PanD., GuoY., HongR. and WangM., Image detail enhancement with spatially guided filters, Signal Process120 (2016), 789–796.
20.
LiS., KangX. and HuJ., Image fusion with guided filtering, IEEE Trans Image Processing22 (2013), 2864–2875.
21.
ZhengC., TanZ.H., PengR. and LiX., Guided spectrogram filtering for speech dereverberation, Acoust134 (2018), 154–159.
22.
BuadesA., CollB. and MorelJ.M., A non-local algorithm for image denoising. In Proceedings of 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Diego, CA, USA, 2005.
23.
JiangB., MengH.Q., MaX.L., WangL., ZhouY., et al., Nighttime image Dehazing with modified models of color transfer and guided image filter, Tools Appl77 (2018), 3125–3141.
24.
DraperN.R. and SmithH., The general regression situation, Applied Regression Analysis, 2014.
25.
HastieT., TibshiraniR. and FriedmanJ., Linear Methods for Regression. In The Elements of Statistical Learning; Springer: Berlin, Germany, 2009, pp. 43–99.
26.
ChaiL., DuJ., LiuQ.F., et al., Using generalized gaussian distributions to improve regression error modeling for deep learning-based speech enhancement. IEEE Trans Audio, Speech, Language Process 2019, pp. 1–1.
27.
WangM., ZhangE. and TangZ., Speech enhancement based on NMF under electric vehicle noise condition, IEEE Access6 (2018), 9147–9159.
28.
VargaA. and SteenekenH.J.M., Assessment for automatic speech recognition: II. NOISEX-92: a database and an experiment to study the effect of additive noise on speech recognition systems, Speech Commun31 (2014), 11–20.
29.
CaoL., ZhangT.Q., GaoH.X., et al., Multi-band spectral subtraction method combined with auditory masking properties for speech enhancement, International Congress on Image & Signal Processing, Chongqing, China, 2013.
30.
TaoZ., ZhaoH. and GongC., Speech enhancement based on masking properties of human auditory system and bark wavelet transform, Chin J Acoust30 (2005), 367–372.
31.
LynnM.M., SuC. and MawK.K., Efficient Feature Extraction for Emotion Recognition System. Forth International Conference for Convergence in Technology, Ujire, India, 2018.
32.
SurendranS. and KumarT.K., Variance Normalized Perceptual Subspace Speech Enhancement, AEU –International Journal of Electronics and Communications74 (2017), 44–54.
33.
PlapousC., MarroC. and ScalartP., Improved signal-to-noise ratio estimation for speech enhancement, IEEE Trans Audio, Speech, Language Process14 (2006), 2098–2108.
34.
ZhengC., ZhouY. and LiX., A modified a priori SNR estimator based on the united speech presence probabilities, J Electron Inf Technol30 (2008), 1680–1683.
35.
LinB.S., WuH.D., et al., Wheeze recognition based on 2D bilateral filtering of spectrogram, Biomedical Engineering Applications Basis and Communications18 (2006), 128–137.
36.
WangJ. and YanL., Bilateral spectrogram filtering for speech denoising. In Proceedings of the National Acoustics Congress, Beijing, China, 2018.
37.
LoizouP.C., Speech enhancement: Theory and practice. Crc Press, 2017.
38.
BhatG.S., ShankarN., ReddyC.K.A. and PanahiI.M.S., A real-Time convolutional neural network based speech enhancement for hearing impaired listeners using smartphone, IEEE Access7 (2019), 78421–78433.
39.
RamR. and MohantyM.N., Performance analysis of adaptive variational mode decomposition approach for speech enhancement, International Journal of Speech Technology21 (2018), 369–381.
40.
ShanD., ZhangX., ZhangC. and LiL., A novel encoder-decoder model via NS-LSTM used for bone-conducted speech enhancement, IEEE Access6 (2018), 62638–62644.
41.
MaJ., HuY. and LoizouP.C., Objective measures for predicting speech intelligibility in noisy conditions based on new band-importance functions, J Acoust Soc Am125 (2009), 3387–3405.
42.
ZhengC., YangJ., ZhangX., SunM. and CaoT., A spectra-based equalization-generation combined framework for throat microphone speech enhancement, IEEE Access6 (2018), 71455–71463.
43.
TaalC.H., HendriksR.C., HeusdensR. and JensenJ., An algorithm for intelligibility prediction of time–frequency weighted noisy speech, IEEE Trans Audio, Speech, Language Process19 (2011), 2125–2136.
