Speech Enhancement using Multiple Transducers

<p>In this thesis, three methods of speech enhancement techniques are investigated with applications in extreme noise environments. Various beamforming techniques are evaluated for their performance characteristics in terms of signal to (distant) noise ratio and tolerance to design imperfections. Two suitable designs are identified with contrasting performance characteristics — the second order differential array, with excellent noise rejection but poor robustness; and a least squares design, with adequate noise rejection and good robustness. Adaptive filters are introduced in the context of a simple noise canceller and later a post-processor for a dual beamformer system. Modifications to the least mean squares (LMS) filter are introduced to tolerate cross-talk between microphones or beamformer outputs. An adaptive filter based post-processor beamforming system is designed and evaluated using a simulation involving speech in noisy environments. The beamforming methods developed are combined with the modified LMS adaptive filter to further reduce noise (if possible) based on correlations between noise signals in a beamformer directed to the talker and a complementary beamformer (nullformer) directed away from the talker. This system shows small, but not insignificant, improvements in noise reduction over purely beamforming based methods. Blind source separation is introduced briefly as a potential future method for enhancing speech in noisy environments. The FastICA algorithm is evaluated on existing data sets and found to perform similarly to the post-processing system developed in this thesis. Future avenues of research in this field are highlighted.</p>

Speech Enhancement using Multiple Transducers

<p>In this thesis, three methods of speech enhancement techniques are investigated with applications in extreme noise environments. Various beamforming techniques are evaluated for their performance characteristics in terms of signal to (distant) noise ratio and tolerance to design imperfections. Two suitable designs are identified with contrasting performance characteristics — the second order differential array, with excellent noise rejection but poor robustness; and a least squares design, with adequate noise rejection and good robustness. Adaptive filters are introduced in the context of a simple noise canceller and later a post-processor for a dual beamformer system. Modifications to the least mean squares (LMS) filter are introduced to tolerate cross-talk between microphones or beamformer outputs. An adaptive filter based post-processor beamforming system is designed and evaluated using a simulation involving speech in noisy environments. The beamforming methods developed are combined with the modified LMS adaptive filter to further reduce noise (if possible) based on correlations between noise signals in a beamformer directed to the talker and a complementary beamformer (nullformer) directed away from the talker. This system shows small, but not insignificant, improvements in noise reduction over purely beamforming based methods. Blind source separation is introduced briefly as a potential future method for enhancing speech in noisy environments. The FastICA algorithm is evaluated on existing data sets and found to perform similarly to the post-processing system developed in this thesis. Future avenues of research in this field are highlighted.</p>

Mechanical Compound Fault Analysis Method Based on Shift Invariant Dictionary Learning and Improved FastICA Algorithm

For mechanical compound fault, it is of great significance to employ the vibration signal of a single-channel compound fault to analyze and realize the separation of multiple fault sources, which is essentially the problem of single-channel blind source separation. Shift invariant K-means singular value decomposition (shift invariant K-SVD) dictionary learning is suitable to extract the periodic and repeated fault features of a rotating machinery fault, hence in this article a single-channel compound fault analysis method is put forward which combines shift invariant K-SVD with improved fast independent component analysis (improved FastICA) algorithm. Firstly, based on single-channel compound fault signal, the shift invariant K-SVD algorithm can be used for learning multiple latent components that can be constructed as a virtual multi-channel signal. Then the improved FastICA algorithm is utilized to realize the separation of multiple fault source signals. With regard to the FastICA algorithm, the third-order convergence Newton iteration method is adopted to improve convergence speed. Moreover, in order to address the problem that FastICA is very sensitive to initialization, a steepest descent method can be applied. The experimental analysis of the compound fault of rolling bearing verifies that the presented method is effective to separate multiple fault source signals and the improved FastICA algorithm can increase convergence rate and overcome the problem of sensitivity to initialization.

A New Robust Decoding Technique of Four Transmitters MIMO STBC System based on FastICA Algorithm

A new robust decoding technique which designed of Multiple-Input Multiple-Output Space–Time Block Code (MIMO-STBC) using Fast Independent Component Analysis (Fast-ICA) based on proposed mixing model has been performed in this paper. This decoding technique is characterized by i) complexity is very low, ii) the speed is high and iii) BER performance is excellent. It can be achieved with any MIMO STBC system with a fewer pilot symbols number. Also, it is reduced decoding time into 1/8 by innovating a simple strategy referred by one source extraction method. Also, this paper includes suitable initializing for the de-mixing vector to solve the ambiguities problem of sign and source of blind source separation (BSS). To test the proposed technique, four transmitters (4Tx) STBC MIMO system was implemented using MATLAB2018. It also found that excellent BER performance associated with a high number of symbols per frame (about 8012 symbols). The simulation results show that the new decoder works for any number of receiver antenna (Nr = 2, 4 and 5). As compare with classical decoding algorithm, it is found that the new decoder provides coding gain (at BER =10-6 ) equal to 1 dB,1.45 dB and 1.76 dB when Nr = 2,4 and 8 respectively, using only 2~3 iterations only.

A novel FastICA algorithm based on improved secant method for Intelligent drive

Blind Source Separation(BSS) is one of the research hotspots in the field of signal processing. In order to improve the accuracy of speech recognition in driving environment, the driver’s speech signal must be enhanced to improve its signal to noise ratio(SNR). Independent component analysis (ICA) algorithm is the most classical and efficient blind statistical signal processing technique. Compared with other improved ICA algorithms, fixed-point algorithm (FastICA) is well known for its fast convergence speed and good robustness. However, the convergence of FastICA algorithm is comparatively susceptible to the initial value selection of the original demixing matrix and the calculation of the iterative process is relatively large. In this paper, the gradient descent method is used to reduce the effect of initial value. What’s more, the improved secant method is proposed to speed up the convergence rate and reduce the amount of computation. As the results of mixed speech separation experiment turn out, the improved algorithm is of better performance relative to the standard FastICA algorithm. Experimental results show that the proposed algorithm improves the speech quality of the target driver. It is suitable for speech separation in driving environment with low SNR.

Orthogonal Approach to Independent Component Analysis Using Quaternionic Factorization

Independent component analysis (ICA) is a popular technique for demixing multi-channel data. The performance of a typical ICA algorithm strongly depends on the presence of additive noise, the actual distribution of source signals, and the estimated number of non-Gaussian components. Often a linear mixing model is assumed and source signals are extracted by data whitening followed by a sequence of plane (Jacobi) rotations. In this article, we develop a novel algorithm, based on the quaternionic factorization of rotation matrices and the Newton-Raphson iterative scheme. Unlike conventional rotational techniques such as the JADE algorithm, our method exploits $4 \times 4$ rotation matrices and uses approximate negentropy as a contrast function. Consequently, the proposed method can be adjusted to a given data distribution (e.g. super-Gaussians) by selecting a suitable non-linear function that approximates the negentropy. Compared to the widely-used, the symmetric FastICA algorithm, the proposed method does not require an orthogonalization step and is more accurate in the presence of multiple Gaussian sources.