A generalized Stein's estimation approach to speech enhancement based on perceptual criteria

We address the problem of speech enhancement using a risk- estimation approach. In particular, we propose the use the Stein’s unbiased risk estimator (SURE) for solving the problem. The need for a suitable finite-sample risk estimator arises because the actual risks invariably depend on the unknown ground truth. We consider the popular mean-squared error (MSE) criterion first, and then compare it against the perceptually-motivated Itakura-Saito (IS) distortion, by deriving unbiased estimators of the corresponding risks. We use a generalized SURE (GSURE) development, recently proposed by Eldar for MSE. We consider dependent observation models from the exponential family with an additive noise model,and derive an unbiased estimator for the risk corresponding to the IS distortion, which is non-quadratic. This serves to address the speech enhancement problem in a more general setting. Experimental results illustrate that the IS metric is efficient in suppressing musical noise, which affects the MSE-enhanced speech. However, in terms of global signal-to-noise ratio (SNR), the minimum MSE solution gives better results.

Compressed acquisition of correlated signals

Distributed compressed sensing exploits information redundancy, inbuilt in multi-signal ensembles with interas well as intra-signal correlations, to reconstruct undersampled signals. In this paper we revisit this problem, albeit from a different perspective, of taking streaming data, from several correlated sources, as input to a real time system which, without any a priori information, incrementally learns and admits each source into the system.

Sub-band envelope approach to obtain instants of significant excitation in speech

In this paper, we propose a new sub-band approach to estimate the glottal activity. The method is based on the spectral harmonicity and the sub-band temporal properties of voiced speech. We propose a method to represent glottal excitation signal using sub-band temporal envelope. Instants of maximum glottal excitation or Glottal Closure Instants (GCI) are extracted from the estimated glottal excitation pattern and the result is compared with a standard GCI computation method, DYPSA [1]. The performance of the algorithm is also compared for the noisy signal and it is shown that the proposed method is less variant to GCI estimation under noisy conditions compared to DYPSA. The algorithm is evaluated on the CMU-ARCTIC database.

Fusion of algorithms for compressed sensing

For compressed sensing (CS), we develop a new scheme inspired by data fusion principles. In the proposed fusion based scheme, several CS reconstruction algorithms participate and they are executed in parallel, independently. The final estimate of the underlying sparse signal is derived by fusing the estimates obtained from the participating algorithms. We theoretically analyze this fusion based scheme and derive sufficient conditions for achieving a better reconstruction performance than any participating algorithm. Through simulations, we show that the proposed scheme has two specific advantages: 1) it provides good performance in a low dimensional measurement regime, and 2) it can deal with different statistical natures of the underlying sparse signals. The experimental results on real ECG signals shows that the proposed scheme demands fewer CS measurements for an approximate sparse signal reconstruction.

Joint pitch-analysis formant-synthesis framework for CS recovery of speech

A joint analysis-synthesis framework is developed for the compressive sensing (CS) recovery of speech signals. The signal is assumed to be sparse in the residual domain with the linear prediction filter used as the sparse transformation. Importantly this transform is not known apriori, since estimating the predictor filter requires the knowledge of the signal. Two prediction filters, one comb filter for pitch and another all pole formant filter are needed to induce maximum sparsity. An iterative method is proposed for the estimation of both the prediction filters and the signal itself. Formant prediction filter is used as the synthesis transform, while the pitch filter is used to model the periodicity in the residual excitation signal, in the analysis mode. Significant improvement in the LLR measure is seen over the previously reported formant filter estimation.

Automatic speech segmentation using probabilistic latent component modeling

Latent variable methods, such as PLCA (Probabilistic Latent Component Analysis) have been successfully used for analysis of non-negative signal representations. In this paper, we formulate PLCS (Probabilistic Latent Component Segmentation), which models each time frame of a spectrogram as a spectral distribution. Given the signal spectrogram, the segmentation boundaries are estimated using a maximum-likelihood approach. For an efficient solution, the algorithm imposes a hard constraint that each segment is modelled by a single latent component. The hard constraint facilitates the solution of ML boundary estimation using dynamic programming. The PLCS framework does not impose a parametric assumption unlike earlier ML segmentation techniques. PLCS can be naturally extended to model coarticulation between successive phones. Experiments on the TIMIT corpus show that the proposed technique is promising compared to most state of the art speech segmentation algorithms.

Detection of the closure-burst transitions of stops and affricates in continuous speech using the plosion index

Automatic and accurate detection of the closure-burst transition events of stops and affricates serves many applications in speech processing. A temporal measure named the plosion index is proposed to detect such events, which are characterized by an abrupt increase in energy. Using the maxima of the pitch-synchronous normalized cross correlation as an additional temporal feature, a rule-based algorithm is designed that aims at selecting only those events associated with the closure-burst transitions of stops and affricates. The performance of the algorithm, characterized by receiver operating characteristic curves and temporal accuracy, is evaluated using the labeled closure-burst transitions of stops and affricates of the entire TIMIT test and training databases. The robustness of the algorithm is studied with respect to global white and babble noise as well as local noise using the TIMIT test set and on telephone quality speech using the NTIMIT test set. For these experiments, the proposed algorithm, which does not require explicit statistical training and is based on two one-dimensional temporal measures, gives a performance comparable to or better than the state-of-the-art methods. In addition, to test the scalability, the algorithm is applied on the Buckeye conversational speech corpus and databases of two Indian languages. (C) 2014 Acoustical Society of America.

RIESZ-TRANSFORM-BASED DEMODULATION OF NARROWBAND SPECTROGRAMS OF VOICED SPEECH

Narrowband spectrograms of voiced speech can be modeled as an outcome of two-dimensional (2-D) modulation process. In this paper, we develop a demodulation algorithm to estimate the 2-D amplitude modulation (AM) and carrier of a given spectrogram patch. The demodulation algorithm is based on the Riesz transform, which is a unitary, shift-invariant operator and is obtained as a 2-D extension of the well known 1-D Hilbert transform operator. Existing methods for spectrogram demodulation rely on extension of sinusoidal demodulation method from the communications literature and require precise estimate of the 2-D carrier. On the other hand, the proposed method based on Riesz transform does not require a carrier estimate. The proposed method and the sinusoidal demodulation scheme are tested on real speech data. Experimental results show that the demodulated AM and carrier from Riesz demodulation represent the spectrogram patch more accurately compared with those obtained using the sinusoidal demodulation. The signal-to-reconstruction error ratio was found to be about 2 to 6 dB higher in case of the proposed demodulation approach.

Progressive fusion of reconstruction algorithms for low latency applications in compressed sensing

Recently, it has been shown that fusion of the estimates of a set of sparse recovery algorithms result in an estimate better than the best estimate in the set, especially when the number of measurements is very limited. Though these schemes provide better sparse signal recovery performance, the higher computational requirement makes it less attractive for low latency applications. To alleviate this drawback, in this paper, we develop a progressive fusion based scheme for low latency applications in compressed sensing. In progressive fusion, the estimates of the participating algorithms are fused progressively according to the availability of estimates. The availability of estimates depends on computational complexity of the participating algorithms, in turn on their latency requirement. Unlike the other fusion algorithms, the proposed progressive fusion algorithm provides quick interim results and successive refinements during the fusion process, which is highly desirable in low latency applications. We analyse the developed scheme by providing sufficient conditions for improvement of CS reconstruction quality and show the practical efficacy by numerical experiments using synthetic and real-world data. (C) 2013 Elsevier B.V. All rights reserved.

Co-registration of speech production datasets from electromagnetic articulography and real-time magnetic resonance imaging

This paper describes a spatio-temporal registration approach for speech articulation data obtained from electromagnetic articulography (EMA) and real-time Magnetic Resonance Imaging (rtMRI). This is motivated by the potential for combining the complementary advantages of both types of data. The registration method is validated on EMA and rtMRI datasets obtained at different times, but using the same stimuli. The aligned corpus offers the advantages of high temporal resolution (from EMA) and a complete mid-sagittal view (from rtMRI). The co-registration also yields optimum placement of EMA sensors as articulatory landmarks on the magnetic resonance images, thus providing richer spatio-temporal information about articulatory dynamics. (C) 2014 Acoustical Society of America

A Committee Machine Approach for Compressed Sensing Signal Reconstruction

Although many sparse recovery algorithms have been proposed recently in compressed sensing (CS), it is well known that the performance of any sparse recovery algorithm depends on many parameters like dimension of the sparse signal, level of sparsity, and measurement noise power. It has been observed that a satisfactory performance of the sparse recovery algorithms requires a minimum number of measurements. This minimum number is different for different algorithms. In many applications, the number of measurements is unlikely to meet this requirement and any scheme to improve performance with fewer measurements is of significant interest in CS. Empirically, it has also been observed that the performance of the sparse recovery algorithms also depends on the underlying statistical distribution of the nonzero elements of the signal, which may not be known a priori in practice. Interestingly, it can be observed that the performance degradation of the sparse recovery algorithms in these cases does not always imply a complete failure. In this paper, we study this scenario and show that by fusing the estimates of multiple sparse recovery algorithms, which work with different principles, we can improve the sparse signal recovery. We present the theoretical analysis to derive sufficient conditions for performance improvement of the proposed schemes. We demonstrate the advantage of the proposed methods through numerical simulations for both synthetic and real signals.

Gammatone Wavelet Cepstral Coefficients for Robust Speech Recognition

We develop noise robust features using Gammatone wavelets derived from the popular Gammatone functions. These wavelets incorporate the characteristics of human peripheral auditory systems, in particular the spatially-varying frequency response of the basilar membrane. We refer to the new features as Gammatone Wavelet Cepstral Coefficients (GWCC). The procedure involved in extracting GWCC from a speech signal is similar to that of the conventional Mel-Frequency Cepstral Coefficients (MFCC) technique, with the difference being in the type of filterbank used. We replace the conventional mel filterbank in MFCC with a Gammatone wavelet filterbank, which we construct using Gammatone wavelets. We also explore the effect of Gammatone filterbank based features (Gammatone Cepstral Coefficients (GCC)) for robust speech recognition. On AURORA 2 database, a comparison of GWCCs and GCCs with MFCCs shows that Gammatone based features yield a better recognition performance at low SNRs.

RAPID HUMAN ACTION RECOGNITION IN H.264/AVC COMPRESSED DOMAIN FOR VIDEO SURVEILLANCE

This paper discusses a novel high-speed approach for human action recognition in H. 264/AVC compressed domain. The proposed algorithm utilizes cues from quantization parameters and motion vectors extracted from the compressed video sequence for feature extraction and further classification using Support Vector Machines (SVM). The ultimate goal of our work is to portray a much faster algorithm than pixel domain counterparts, with comparable accuracy, utilizing only the sparse information from compressed video. Partial decoding rules out the complexity of full decoding, and minimizes computational load and memory usage, which can effect in reduced hardware utilization and fast recognition results. The proposed approach can handle illumination changes, scale, and appearance variations, and is robust in outdoor as well as indoor testing scenarios. We have tested our method on two benchmark action datasets and achieved more than 85% accuracy. The proposed algorithm classifies actions with speed (>2000 fps) approximately 100 times more than existing state-of-the-art pixel-domain algorithms.

Estimation of voice-onset time in continuous speech using temporal measures

This paper proposes an automatic acoustic-phonetic method for estimating voice-onset time of stops. This method requires neither transcription of the utterance nor training of a classifier. It makes use of the plosion index for the automatic detection of burst onsets of stops. Having detected the burst onset, the onset of the voicing following the burst is detected using the epochal information and a temporal measure named the maximum weighted inner product. For validation, several experiments are carried out on the entire TIMIT database and two of the CMU Arctic corpora. The performance of the proposed method compares well with three state-of-the-art techniques. (C) 2014 Acoustical Society of America