This paper presents a new method for speech enhancement based on a dictionary learning method. The proposed approach is based on using coherence measure in dictionary learning. Data required for better fitting to atoms in sparse representation of noise is provided by a noise estimation algorithm that causes noise dictionary to be trained with the same data size as speech signal. To decrease coherence between dictionaries after the training step, a new method is applied to yield incoherent dictionaries. In sparse representation of speech data, the highest energy atoms of noise dictionary are replaced with the lowest energy atoms, under certain conditions. A similar replacement can happen in sparse representation of noise data. Furthermore, in this paper, only one noise dictionary, chosen by a classification method, is used in speech enhancement step, resulting in a faster algorithm. Objective and subjective measures are used for evaluating the simulation results. According to experimental results, the proposed algorithm has been found superior in performance and computation overhead in comparison with the earlier methods in this context. Moreover, this method achieves significantly better results compared with baseline methods such as multi‐band and geometric spectral subtraction.

The aim of this paper is to introduce a novel phase-based feature representation for robust speech recognition. This method consists of four main parts: autoregressive (AR) model extraction, group delay function (GDF) computation, compression, and scale information augmentation. Coupling GDF with an AR model results in a high-resolution estimate of the power spectrum with low frequency leakage. The compression step includes two stages similar to MFCC without taking a logarithm of the output energies. The fourth part augments the phase-based feature vector with scale information which is based on the Hilbert transform relations and complements the phase spectrum information. In the presence of additive and convolutional noises, the proposed method has led to 15% and 12% reductions in the averaged error rates, respectively (SNR ranging from 0 to 20 dB), compared to the standard MFCCs.

Speech segmentation to covariance‐stationary regions is of great interest in various fields of speech processing. This study presents a novel sequential segmentation algorithm relying on Random Matrix Theory (RMT). The proposed approach utilises an RMT‐derived test statistic checking stationarity of sample covariance matrices. Utilising the statistical properties of the sample eigenvalues recently investigated in the RMT literature, a new expression for the decision threshold is derived in the non‐asymptotic case where the number of samples is comparable with the dimension. This derivation is of great help to detect very short non‐stationary intervals where the performance of most segmentation methods based on Autoregressive models degrades. Furthermore, based on the results of previous step, a new segmentation procedure is proposed. This procedure is applied to both synthetic and real‐world speech data. Comparing the obtained simulation results with the state‐of‐the‐art segmentation algorithms, the proposed approach has demonstrated good performance with extremely lower computational cost.

Multichannel blind deconvolution (MCBD) algorithms are known to suffer from an extensive computational complexity problem, which makes them impractical for blind source separation (BSS) of speech and audio signals. This problem is even more serious with noncausal MCBD algorithms that must be used in many frequently occurring BSS setups. In this paper, we propose a novel frequency domain algorithm for the efficient implementation of noncausal multichannel blind deconvolution. A block-wise formulation is first developed for filtering and adaptation of filter coefficients. Based on this formulation, we present a modified overlap-save procedure for noncausal filtering in the frequency domain. We also derive update equations for training both causal and anti-causal filters in the frequency domain. Our evaluations indicate that the proposed frequency domain implementation reduces the computational requirements of the algorithm by a factor of more than 100 for typical filter lengths used in blind speech separation. The algorithm is employed successfully for the separation of speech mixtures in a reverberant room. Simulation results demonstrate the superior performance of the proposed algorithm over causal MCBD algorithms in many potential source and microphone positions. It is shown that in BSS problems, causal MCBD algorithms with center-spike initialization do not always converge to a delayed form of the desired noncausal solution, further revealing the need for an efficient noncausal MCBD algorithm.

In this paper we present a novel feature extraction algorithm based on group delay function for robust speech recognition. The modified group delay function (MODGDF) is the main feature extraction method based on group delay function, generally used for robust speech recognition. The recognition tests indicate this feature does not provide notably better results in the presence of additive noise in comparison with MFCC. In the presence of convolutional noise, the performance of MODGDF is considerably lower than MFCC. The method proposed in this paper is simple and makes more efficient utilization of the high resolution property of GDF. It is formed from three main parts which are signal modeling, GDF computation based on extracted model, and compression. The recognition results obtained over AURORA 2.0 task indicate its superior performance in comparison with MODGDF and MFCC.

A robust image watermarking scheme in the ridgelet transform domain is proposed in this paper. Due to the use of the ridgelet domain, sparse representation of an image which deals with line singularities is obtained. In order to achieve more robustness and transparency, the watermark data is embedded in selected blocks of the host image by modifying the amplitude of the ridgelet coefficients which represent the most energetic direction. Since the probability distribution function of the ridgelet coefficients is not known, we propose a universally optimum decoder to perform the watermark extraction in a distribution-independent fashion. Decoder extracts the watermark data using the variance of the ridgelet coefficients of the most energetic direction in each block. Furthermore, since the decoder needs the noise variance to perform decoding, a robust noise estimation scheme is proposed. Moreover, the implementation of error correction codes on the proposed method is investigated. Analytical derivation of bit error probability is also carried out and experimental results prove its accuracy. Simulation also shows outstanding robustness of the proposed scheme against common attacks, especially additive white noise and JPEG compression.

In this paper, a novel arrangement for quantizer levels in the Quantization Index Modulation (QIM) method is proposed. Due to perceptual advantages of logarithmic quantization, and in order to solve the problems of a previous logarithmic quantization-based method, we used the compression function of mu-Law standard for quantization. In this regard, the host signal is first transformed into the logarithmic domain using the mu-Law compression function. Then, the transformed data is quantized uniformly and the result is transformed back to the original domain using the inverse function. The scalar method is then extended to vector quantization. For this, the magnitude of each host vector is quantized on the surface of hyperspheres which follow logarithmic radii. Optimum parameter mu for both scalar and vector cases is calculated according to the host signal distribution. Moreover, inclusion of a secret key in the proposed method, similar to the dither modulation in QIM, is introduced. Performance of the proposed method in both cases is analyzed and the analytical derivations are verified through extensive simulations on artificial signals. The method is also simulated on real images and its performance is compared with previous scalar and vector quantization-based methods. Results show that this method features stronger a watermark in comparison with conventional QIM and, as a result, has better performance while it does not suffer from the drawbacks of a previously proposed logarithmic quantization algorithm.

In this paper, a new method for logarithmic quantization index modulation (QIM) is proposed. In this regard a logarithmic function is first applied to the host signal. Then the transformed signal is quantized using uniform quantization as conventional QIM to embed watermark data within. Finally using inverse transform the watermarked signal is obtained. The watermark extraction is performed using minimum distance decoder. The optimum parameter for data embedding with minimum quantization distortion is derived. Also the probability of error is analytically calculated and verified by simulation. Furthermore data hiding using secret key is proposed and the probability of error is obtained. Simulation results show that the proposed method outperforms the conventional QIM in terms of robustness when the perceptual quality of watermarked image for both methods are similar. Moreover, simulation shows that the proposed scheme has outstanding robustness in comparison with a recent quantization based data hiding method.

This paper presents a Multiplicative Patchwork Method (MPM) for audio watermarking. The watermark signal is embedded by selecting two subsets of the host signal features and modifying one subset multiplicatively regarding the watermark data, whereas another subset is left unchanged. The method is implemented in wavelet domain and approximation coefficients are used to embed data. In order to have an error-free detection, the watermark data is inserted only in the frames where the ratio of the energy of subsets is between two predefined values. Also in order to control the inaudibility of watermark insertion, we use an iterative algorithm to reach a desired quality for the watermarked audio signal. The quality of watermarked signal is evaluated in each iteration using Perceptual Evaluation of Audio Quality (PEAQ) method. The probability of error is also derived for the watermarking scheme and simulation results prove the validity of the analytical derivations. Simulation results show that MPM is robust against various common attacks such as noise addition, filtering, echo, MP3 compression, etc. In comparison to the original patchwork method and its modified versions, and some recent methods, MPM provides more robustness and inaudibility of the watermark insertion.

Nowadays, the fuzzy c-means method (FCM) became one of the most popular clustering methods based on minimization of a criterion function. However, the performance of this clustering algorithm may be significantly degraded in the presence of noise. This paper presents a robust clustering algorithm called robust weighted fuzzy c-means (RWFCM). We used a new objective function that uses some kinds of weights for reducing the infection of noises in clustering. Experimental results show that compared to three well-known clustering algorithms, namely, the fuzzy possibilistic c-means (FPCM), credibilistic fuzzy c-means (CFCM) and density weighted fuzzy c-means (DWFCM), RWFCM is less sensitive to outlier and noise and has an acceptable computational complexity.

In this paper, a new acoustic model called time-inhomogeneous hidden Bernoulli model (TI-HBM) is introduced as an alternative to hidden Markov model (HMM) in automatic speech recognition. Contrary to HMM, the state transition process in TI-HBM is not a Markov process; rather it is an independent (generalized Bernoulli) process. This difference leads to elimination of dynamic programming at state-level in TI-HBM decoding process. Thus, the computational complexity of TI-HBM for probability evaluation and state estimation is O(NL) (instead of O(N2L) in the HMM case). As a new framework for phone duration modeling, TI-HBM is able to model acoustic-unit duration (e.g. phone duration) by using a built-in parameter named survival probability. Similar to the HMM case, three essential problems in TI-HBM have been solved. An EM-algorithm based method has been proposed for training TI-HBM parameters. Experiments in phone recognition for Persian (Farsi) spoken language show that the TI-HBM has some advantages over HMM (e.g., more simplicity and increased speed in recognition phase), and also outperforms HMM in terms of phone recognition accuracy.

In this paper, a digital watermarking system based on vector quantization is presented. Each vector containing N samples is mapped on the surface of the hyperspheres each of which are associated with a message to embed the digital watermark. We called this method vector quantization index modulation (VQIM) since it is conventional QIM in the N-dimensional space. The performance of the method and its comparison to orthogonal code-based watermarking is investigated. Furthermore, we implemented the VQIM method on a real audio watermarking system and adopted it with the human auditory system. The experimental results show the robustness of this scheme against common attacks in audio watermarking such as MP3 compression, lowpass filtering, resampling etc.

In this paper, we propose a novel detector for multiplicative watermarking. The decoder extracts the watermark data by comparing the variances of the watermarked signal and the original signal. Due to the use of the variance test, the decoder works independent of the distribution of the host signal. This is the major advantage of this decoder per other decoders. The decoder is optimized under the Additive White Gaussian Noise channel by calculating the best threshold value. In order to show the optimal performance of this decoder on audio signals, a Maximum likelihood (ML) decoder is also introduced by considering a Gaussian distribution for the host signal. Furthermore, PEAQ algorithm is used for controlling the inaudibility of the watermark data insertion. Using this algorithm, the watermark strength factor updates automatically every 200 ms, according to the quality which is desired for the watermarked signal. Simulation results showed that the proposed decoder is extremely robust to the common audio watermarking attacks and slightly better than the ML decoder under Additive White Gaussian Noise attack.

The duty of speaker diarization comprises of answering the question ldquoWho spoke when?rdquo. In this paper, we present an approach comprising of PSO (particle swarm optimization) algorithm, which encodes possible segmentations of an audio record by measuring mutual information between these segments and the audio data.. This measure is used as the fitness function for the PSO. This algorithm has been tested on two actual sets of data with up to 8 speakers for speaker diarization, and has led to very good results in all test problems. The results have been compared to the same approach using genetic algorithm (GA) and the widely used DISTBIC algorithm in several practical situations, and found to be superior in most of the cases. No assumptions have been made about prior knowledge of speech signal characteristics. However, we assume that the speakers do not speak simultaneously and that we have no real-time constraints.

This paper presents a new front-end for robust speech recognition. Two scenarios are used for the features extracted in autocorrelation and group delay domains. These new front-end scenarios focuses on the spectral peaks of speech in two mentioned domains. Therefore we address the issue of using spectral peak location information in a feature vector for robust speech recognition. A task of speaker-independent isolated-word recognition was used to demonstrate the efficiency of these robust front-end diagrams. The cases of white noise and different colored noises such as babble, factory and car noises were tested. Experimental results show significant improvements in comparison to the results obtained using traditional front-end diagrams

Stress is known as an important prosodic property of speech. Recognition of stressed speech, due to the effects of stress on acoustic features of speech, has always been considered more difficult, in comparison to unstressed speech. On the other hand, preparing a large speech corpus to include all different possible occasions of stressed speech, for the purpose of training in speech recognition, is very difficult. This is true for any language and Farsi (Persian) is no exception. Due to the importance of formant frequencies, they have long been considered in the recognition task. In this paper, we report on our efforts to find out the potential effects of stress on the formant frequencies. The results of our statistical evaluations demonstrate changes in the formant frequencies in certain directions, due to stress, that may be found helpful in improving automatic recognition of stressed speech.

The log energy parameter has long been used as an extension to the basic cepstral feature vector in speech recognition. The use of a normalization technique for the log energy parameter has also been widely accepted. In this paper, a simple energy normalization scheme is introduced that allows direct use of the frame energy parameter in speech recognition and performs well in the presence of noise. Its combination with traditional cepstral mean and variance normalizations has led to error rate improvements of up to 55% on the Aurora 2 task, in comparison to the baseline clean-trained system using feature set including the log energy parameter. This achievement has been obtained with neither complicated programming nor computation expensive routines. The performance of this scheme on an utterance-wide basis has been close to that of the off-line speaker-wide normalization, which makes it a good candidate for practical systems.

Stress plays a crucial role in the understanding of speech by human listeners. However, automatic speech recognition results deteriorate in the presence of stress due to the change it causes in the speech parameters. Meanwhile, due to the vast diversity of the presence of stress in speech, a speech corpus that contains the majority of different stress conditions is difficult to obtain in real world. Therefore, other ways to improve stressed speech recognition performance have to be taken into account. In previous works, we have evaluated the effects of stress on several speech parameters such as phone durations, pitch and formant frequencies. In this paper, the use of formants in stressed speech recognition will be discussed. We have found that formants and their dynamics (slopes) are useful in improving speech recognition rates both in stressed and unstressed conditions.

This paper presents a general overview of the current research activities in the European PICASSO project on speaker verification for telephone applications. First, the general formalism used by the project is described. Then the scientific issues under focus are discussed in detail. Finally, the paper briefly describes the Picassoft research platform. Along the article, entry points to more specific work also published in the Eurospeech99 proceedings are given.