Audio coding via EMD

Boudraa, Abdel-Ouahab; Khaldi, Kais; Chonavel, Thierry; Hadj-Alouane, M. Turki; Komaty, Ali

doi:10.1016/j.dsp.2020.102770

Cited by 6 publications

(2 citation statements)

References 31 publications

(38 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A salient property of an IMF is that it can be fully described by its extrema and this can be exploited for different applications such as frequency analysis [1], image and signals coding [35], [36], or signals watermarking [37]. Thus, filter bank structure is quantitatively appreciated by measuring the frequency content of each IMF using its extrema, and more precisely its zero-crossings (ZCs).…”

Section: B Filter Bank Structurementioning

confidence: 99%

On the Behavior of MEMD in Presence of Multivariate Fractional Gaussian Noise

Komaty¹,

Boudraa

Flandrin

et al. 2021

IEEE Trans. Signal Process.

Self Cite

View full text Add to dashboard Cite

Multivariate empirical mode decomposition (MEMD) has been introduced to make standard EMD suitable for direct multichannel signals processing. Unlike EMD, MEMD is able to align sifted intrinsic mode functions (IMFs) from multiple data channels. The aim of this work is to analyze the behavior of MEMD under multivariate fGn (MfGn) with different Hurst exponents, H, and strengths of link between pairs of channels, ρ, of each sifted IMF. We report results supporting the claim that, regardless of ρ values and for both MfGns long-range and short-range dependent, MEMD acts as filter bank on each channel of the input multivariate signal. Whatever the ρ and H values, this equivalent filter bank structure is dyadic with constant-Q band-pass filters. The observed self-similar filter bank structure leads to a deeper statistical studies of the variance distribution and zero-crossings alignment in order to express this self-similarity in terms of spectral densities of multidimensional IMFs. These statistical properties generalize what was previously conducted for EMD to MEMD and estimation strategy of H exponent is proposed. The filter bank behavior of MEMD is illustrated on real turbulent flow data and the estimated H exponent brings out the long-range-dependent nature of the turbulent flow data. An application to EEG data is also proposed.

show abstract

Section: B Filter Bank Structurementioning

confidence: 99%

On the Behavior of MEMD in Presence of Multivariate Fractional Gaussian Noise

Komaty¹,

Boudraa

Flandrin

et al. 2021

IEEE Trans. Signal Process.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Dabbabi Karim et al [14] presented a new method to optimize audio classification and segmentation by utilizing the Genetic Algorithm using Self-Orga-nizing Maps (GASOM) algorithm for their multimedia data set. Audio coding using empirical mode decomposition has been presented in [15]. In this method, the audio signal has been broken down into intrinsic oscillatory components and encoded.…”

Section: Introductionmentioning

confidence: 99%

Robust dual-tone multi-frequency tone detection using k-nearest neighbour classifier for a noisy environment

Maity

Periasamy

Bhargava

2021

ACI

View full text Add to dashboard Cite

PurposeDue to the continuous and rapid evolution of telecommunication equipment, the demand for more efficient and noise-robust detection of dual-tone multi-frequency (DTMF) signals is most significant.Design/methodology/approachA novel machine learning-based approach to detect DTMF tones affected by noise, frequency and time variations by employing the k-nearest neighbour (KNN) algorithm is proposed. The features required for training the proposed KNN classifier are extracted using Goertzel's algorithm that estimates the absolute discrete Fourier transform (DFT) coefficient values for the fundamental DTMF frequencies with or without considering their second harmonic frequencies. The proposed KNN classifier model is configured in four different manners which differ in being trained with or without augmented data, as well as, with or without the inclusion of second harmonic frequency DFT coefficient values as features.FindingsIt is found that the model which is trained using the augmented data set and additionally includes the absolute DFT values of the second harmonic frequency values for the eight fundamental DTMF frequencies as the features, achieved the best performance with a macro classification F1 score of 0.980835, a five-fold stratified cross-validation accuracy of 98.47% and test data set detection accuracy of 98.1053%.Originality/valueThe generated DTMF signal has been classified and detected using the proposed KNN classifier which utilizes the DFT coefficient along with second harmonic frequencies for better classification. Additionally, the proposed KNN classifier has been compared with existing models to ascertain its superiority and proclaim its state-of-the-art performance.

show abstract