Critical band analysis-synthesis

Petersen, T.; Boll, S.

doi:10.1109/icassp.1981.1171198

Cited by 7 publications

(7 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In seismology, this approach to the timefrequency expansion formalized by Dziewonski et al (1969) is known as the multiple filter technique. Similarly, in the context of the short-time Fourier transform, Gambardella (1971) proposes the use of modulated windows of constant Q, instead of constant length, for audio analysis giving a 'form invariant under time scaling', and Petersen & Boll (1983) show how to sample this time-frequency domain efficiently. Later on, the study of these ideas in physics, engineering and pure and applied mathematics converged to the wavelet transform (e.g.…”

Section: Seeking For An Efficient Representation Of the Time-frequencmentioning

confidence: 99%

Extracting surface waves, hum and normal modes: time-scale phase-weighted stack and beyond

Ventosa¹,

Schimmel²,

Stutzmann

2017

Geophysical Journal International

View full text Add to dashboard Cite

Stacks of ambient noise correlations are routinely used to extract empirical Green's functions (EGFs) between station pairs. The time-frequency phase-weighted stack (tf-PWS) is a physically intuitive nonlinear denoising method that uses the phase coherence to improve EGF convergence when the performance of conventional linear averaging methods is not sufficient. The high computational cost of a continuous approach to the time-frequency transformation is currently a main limitation in ambient noise studies. We introduce the timescale phaseweighted stack (ts-PWS) as an alternative extension of the phase-weighted stack that uses complex frames of wavelets to build a time-frequency representation that is much more efficient and fast to compute and that preserve the performance and flexibility of the tf-PWS. In addition, we propose two strategies: the unbiased phase coherence and the two-stage ts-PWS methods to further improve noise attenuation, quality of the extracted signals and convergence speed. We demonstrate that these approaches enable to extract minor-and major-arc Rayleigh waves (up to the sixth Rayleigh wave train) from many years of data from the GEOSCOPE global network. Finally we also show that fundamental spheroidal modes can be extracted from these EGF.

show abstract

Section: Seeking For An Efficient Representation Of the Time-frequencmentioning

confidence: 99%

Extracting surface waves, hum and normal modes: time-scale phase-weighted stack and beyond

Ventosa¹,

Schimmel²,

Stutzmann

2017

Geophysical Journal International

View full text Add to dashboard Cite

show abstract

“…In the past two decades, various schemes to implement critical-band analysis [4][5][6][7][8][9][10] have been proposed for speech applications. These methods can be classified into four main approaches: (i) direct digital implementation of the criticalband filterbank, (ii) FFT method, (iii) constant-Q transform (CQT) method, and (iv) wavelet packet transform (WPT) method.…”

Section: Introductionmentioning

confidence: 99%

“…Some typical applications of the FFT method include audio coding [5] and 2 EURASIP Journal on Advances in Signal Processing speech recognition [6]. One of the CQT methods [7] uses constant-Q filters to approximate the critical-band filtering in the high frequency range. In the lower frequency range, the constant-bandwidth coefficients are obtained by summing the constant-Q filters coefficients within each constantbandwidth band in question.…”

Section: Introductionmentioning

confidence: 99%

Efficient Algorithm and Architecture of Critical-Band Transform for Low-Power Speech Applications

Chao

Gan

2007

EURASIP J. Adv. Signal Process.

View full text Add to dashboard Cite

An efficient algorithm and its corresponding VLSI architecture for the critical-band transform (CBT) are developed to approximate the critical-band filtering of the human ear. The CBT consists of a constant-bandwidth transform in the lower frequency range and a Brown constant-Q transform (CQT) in the higher frequency range. The corresponding VLSI architecture is proposed to achieve significant power efficiency by reducing the computational complexity, using pipeline and parallel processing, and applying the supply voltage scaling technique. A 21-band Bark scale CBT processor with a sampling rate of 16 kHz is designed and simulated. Simulation results verify its suitability for performing short-time spectral analysis on speech. It has a better fitting on the human ear critical-band analysis, significantly fewer computations, and therefore is more energy-efficient than other methods. With a 0.35 μm CMOS technology, it calculates a 160-point speech in 4.99 milliseconds at 234 kHz. The power dissipation is 15.6 μW at 1.1 V. It achieves 82.1% power reduction as compared to a benchmark 256-point FFT processor.

show abstract

“…However, as we described in Chapter 2, psychoacoustical studies show that the human ear performs spectral analysis on the acoustic signal in the form of a filterbank with nonuniform critical bandwidths [38]. As shown in Table 2.1, for a wide-band speech signal with a bandwidth of 8 kHz, there are 21 critical bands for the Bark scale described by Zwicker [40] and 24 bands for the Mel scale (ii) FFf method [39], (iii) constant-Q transform (CQT) method [71], [72] and (iv) methods in the next section. Recently, low-power VLSI speech systems, such as Algorithm and Architecture of Critical-Band Transform Chapter 4 make them not suitable for low-power VLSI realization.…”

Section: Discussionmentioning

confidence: 99%

“…One of the CQT-based methods [71] uses constant-Q filters to approximate the critical-band filtering in the high frequency range. In the lower frequency range, the constant-bandwidth coefficients are obtained by summing the constant-Q filters coefficients within each constant-bandwidth band.…”

Section: Constant-q Transform Methodsmentioning

confidence: 99%

Very large scale integration architectures of discrete transforms for digital signal processing applications

Wang¹

View full text Add to dashboard Cite

Luo Jianwen, Associate Professor Jong Ching Chuen, and Associate Professor Chang Chip Hong, for their friendship, help, discussions, suggestions and encouragement.I do also appreciate the support and help from my colleagues in the FTD Solutions Pte Ltd and in the Centre for Signal Processing, NTD. Especially, I would like to thank Qian Yue, Lu Dawei, Yang Fan, and Dr. Yu Zhuliang for their valuable comments, suggestions and criticisms.I have enjoyed the company and friendship of all of my friends in Singapore. I would like to thank them for making the past five years an unforgettable and invaluable experience in my life.Last, I wish to express my deepest gratitude to my dear parents, my dear brother, and my lovely four year old nephew. It is indeed their deep love and consistent encouragement that has kept me focused, determined, cheered up, and never to say give up.

show abstract

Critical band analysis-synthesis

Cited by 7 publications

References 6 publications

Extracting surface waves, hum and normal modes: time-scale phase-weighted stack and beyond

Extracting surface waves, hum and normal modes: time-scale phase-weighted stack and beyond

Efficient Algorithm and Architecture of Critical-Band Transform for Low-Power Speech Applications

Very large scale integration architectures of discrete transforms for digital signal processing applications

Contact Info

Product

Resources

About