New methodology for adaptive vector quantisation

An image sequence coding scheme for low bit-rate video coding is presented. A new methodology is proposed for adaptive vector quantisation (AVQ), where the codebook is updated with new code-vectors. The new code-vectors replace less significant ones in the codebook, based on a novel scoring criterion that utilises a forgetting factor and codebook half-life. The proposed AVQ method gives rise to an additional performance enhancement of approximately 1 dB over frequency scoring techniques     

Regularised nonlinear blind signal separation using sparsely connected network

A nonlinear approach based on the Tikhonov regularised cost function is presented for blind signal separation of nonlinear mixtures. The proposed approach uses a multilayer perceptron as the nonlinear demixer and combines both information theoretic learning and structural complexity learning into a single framework. It is shown that this approach can be jointly used to extract independent components while constraining the overall perceptron network to be as sparse as possible. The update algorithm for the nonlinear demixer is subsequently derived using the new cost function. Sparseness in the network connection is utilised to determine the total number of layers required in the multilayer perceptron and to prevent the nonlinear demixer from outputting arbitrary independent components. Experiments are meticulously conducted to study the performance of the new approach and the outcomes of these studies are critically assessed for performance comparison with existing methods.     

Single channel informed signal separation using artificial‐stereophonic mixtures and exemplar‐guided matrix factor deconvolution

In this paper, a method is proposed to tackle the problem of single channel audio separation. The proposed method leverages on the exemplar source is used to emulate the targeted speech signal. A multicomponent nonnegative matrix factor 2D deconvolution (NMF2D) is proposed to model the temporal and spectral changes and the number of spectral basis of the audio signals. The paper proposes an artificial auxiliary channel to imitate a pair of stereo mixture signals, which is termed as “artificial‐stereophonic mixtures.” The artificial‐stereophonic mixtures and the exemplar source are jointly used to guide the factorization process of the NMF2D. The factorization is adapted under a hybrid framework that combines the generalized expectation–maximization algorithm with multiplicative update adaptation. The proposed algorithm leads to fast and stable convergence and ensures the nonnegativity constraints of the solution are satisfied. Adaptive sparsity has also been introduced on each sparse parameter in the multicomponent NMF2D model when the exemplar deviates from the target signal. Experimental results have shown the competence of the proposed algorithms in comparison with other algorithms.     

Nonlinear blind source separation using a hybrid RBF-FMLP network

A novel scheme for blind source separation of nonlinearly mixed signals is developed using a hybrid system based on radial basis function (RBF) and feedforward multilayer perceptron (FMLP) networks. In this paper, the development of the proposed RBF-FMLP network is discussed, which hinges on the theory of nonlinear regularisation. The proposed network uses simultaneously local and global mapping bases to perform both signal separation and reconstruction of continuous signals in addition to signals that exhibit a high degree of fluctuation. The parameters of the proposed system are estimated jointly using the generalised gradient descent approach thereby rendering the training process relatively simple and efficient in computation. Simulations of both synthetic and speech signals have been undertaken to verify the efficacy of the proposed scheme in terms of speed, accuracy and robustness against noise.     

Neural network approach to blind signal separation of mono-nonlinearly mixed sources

A new result is developed for separating nonlinearly mixed signals in which the nonlinearity is characterized by a class of strictly monotonic continuously differentiable functions. The structure of the blind inverse system is explicitly derived within the framework of maximum likelihood estimation and the system culminates to a special architecture of the 3-layer perceptron neural network where the parameters in the first layer are inversely related to the output layer. The proposed approach exploits both the structural and signal constraints to search for the solution and assumes that the cumulants of the source signals are known a priori. A novel statistical algorithm based on the hybridization of the generalized gradient algorithm and metropolis algorithm has been derived for training the proposed perceptron which results in improved performance in terms of accuracy and convergence speed. Simulations and real-life experiment have also been conducted to verify the efficacy of the proposed scheme in separating the nonlinearly mixed signals.     

Non-linear independent component analysis using series reversion and Weierstrass network

The problem of blind separation of independent sources in non-linear mixtures is considered and the focus of this work is on a new type of non-linear mixture in which a linear mixing matrix is sandwiched between two mutually reverse non-linearities. The demixing system culminates to a novel Weierstrass network that is shown to successfully restore the original source signals under the non-linear mixing conditions. The corresponding parameter learning algorithm for the proposed network is presented through formal mathematical derivation. The authors show for the first time a new result based on the theory of forward series and series reversion that is integrated into a neural network to implement the proposed demixer. Simulations, including both synthetic and recorded signals, have been carried out to verify the efficacy of the proposed method. It is shown that the Weierstrass network outperforms other tested independent component analysis (ICA) methods (linear ICA, radial-basis function and multilayer perceptron network) in terms of speed and accuracy.     

Blind Source Separation of Postnonlinear Convolutive Mixture

In this paper, a novel solution is developed to solve blind source separation of postnonlinear convolutive mixtures. The proposed model extends the conventional linear instantaneous mixture model to include both convolutive mixing and postnonlinear distortion. The maximum-likelihood (ML) approach solution based on the expectation-maximization (EM) algorithm is developed to estimate the source signals and the parameters in the proposed nonlinear model. In the proposed solution, the sufficient statistics associated with the source signals are estimated in the E-step, while the model parameters are optimized through these statistics in the M-step. However, the complication resulted from the postnonlinear function associated with the mixture renders these statistics difficult to be formulated in a closed form and hence causes intractability in the parameter optimization. A computationally efficient algorithm is proposed which uses the extended Kalman smoother (EKS) to facilitate the E-step tractable and a set of self-updated polynomials is used as the nonlinearity estimator to facilitate closed form estimations of the parameters in the M-step. The theoretical foundation of the proposed solution has been rigorously developed and discussed in details. Both simulations and recorded speech signals have been carried out to verify the success and efficacy of the proposed algorithm. Remarkable improvement has been obtained when compared with the existing algorithms.     

Nonlinear signal separation for multinonlinearity constrained mixing model

In this letter, a new type of nonlinear mixture is derived and developed into a multinonlinearity constrained mixing model. The proposed signal separation solution integrates the Theory of Series Reversion with a polynomial neural network whereby the hidden neurons are spanned by a set of mutually reversed activation functions. Simulations have been undertaken to support the theory of the proposed scheme and the results indicate promising performance.