Luz García

Statistical voice activity detection using a multiple observation likelihood ratio test

Currently, there are technology barriers inhibiting speech processing systems that work in extremely noisy conditions from meeting the demands of modern applications. This letter presents a new voice activity detector (VAD) for improving speech detection robustness in noisy environments and the performance of speech recognition systems. The algorithm defines an optimum likelihood ratio test (LRT) involving multiple and independent observations. The so-defined decision rule reports significant improvements in speech/nonspeech discrimination accuracy over existing VAD methods that are defined on a single observation and need empirically tuned hangover mechanisms. The algorithm has an inherent delay that, for several applications, including robust speech recognition, does not represent a serious implementation obstacle. An analysis of the overlap between the distributions of the decision variable shows the improved robustness of the proposed approach by means of a clear reduction of the classification error as the number of observations is increased. The proposed strategy is also compared to different VAD methods, including the G.729, AMR, and AFE standards, as well as recently reported algorithms showing a sustained advantage in speech/nonspeech detection accuracy and speech recognition performance.

Improved Voice Activity Detection Using Contextual Multiple Hypothesis Testing for Robust Speech Recognition

This paper shows an improved statistical test for voice activity detection in noise adverse environments. The method is based on a revised contextual likelihood ratio test (LRT) defined over a multiple observation window. The motivations for revising the original multiple observation LRT (MO-LRT) are found in its artificially added hangover mechanism that exhibits an incorrect behavior under different signal-to-noise ratio (SNR) conditions. The new approach defines a maximum a posteriori (MAP) statistical test in which all the global hypotheses on the multiple observation window containing up to one speech-to-nonspeech or nonspeech-to-speech transitions are considered. Thus, the implicit hangover mechanism artificially added by the original method was not found in the revised method so its design can be further improved. With these and other innovations, the proposed method showed a higher speech/nonspeech discrimination accuracy over a wide range of SNR conditions when compared to the original MO-LRT voice activity detector (VAD). Experiments conducted on the AURORA databases and tasks showed that the revised method yields significant improvements in speech recognition performance over standardized VADs such as ITU T G.729 and ETSI AMR for discontinuous voice transmission and the ETSI AFE for distributed speech recognition (DSR), as well as over recently reported methods.

Parametric Nonlinear Feature Equalization for Robust Speech Recognition

A new front-end normalization algorithm that uses a parametric nonlinear transformation is proposed in this paper. The method improves histogram equalization based nonlinear transformations by finding a simple and computationally inexpensive parametric expression of the nonlinear transformation. The new parametric approach relies on a two Gaussian model for the probability distribution of the features, and on a simple Gaussian classifier to label the input frames as belonging to the speech or non-speech classes. The result is a more robust equalization, less dependent on the percentage of speech and non-speech frames. Recognition experiments on the AURORA 4 database have been performed and the effectiveness of the algorithm is analyzed in comparison with other linear and nonlinear feature equalization techniques

Seismic hydraulic fracture migration originated by successive deep magma pulses: The 2011–2013 seismic series associated to the volcanic activity of El Hierro Island

In this manuscript we present a new interpretation of the seismic series that accompanied eruptive activity off the coast of El Hierro, Canary Islands, during 2011–2013. We estimated temporal variations of the Gutenberg-Richter b value throughout the period of analysis, and performed high-precision relocations of the preeruptive and syneruptive seismicity using a realistic 3-D velocity model. Our results suggest that eruptive activity and the accompanying seismicity were caused by repeated injections of magma from the mantle into the lower crust. These magma pulses occurred within a small and well-defined volume resulting in the emplacement of fresh magma along the crust-mantle boundary underneath El Hierro. We analyzed the distribution of earthquake hypocenters in time and space in order to assess seismic diffusivity in the lower crust. Our results suggest that very high earthquake rates underneath El Hierro represent the response of a stable lower crust to stress perturbations with pulsatory character, linked to the injection of magma from the mantle. Magma input from depth caused large stress perturbations to propagate into the lower crust generating energetic seismic swarms. The absence of any preferential alignment in the spatial pattern of seismicity reinforces our hypothesis that stress perturbation and related seismicity, had diffusive character. We conclude that the temporal and spatial evolution of seismicity was neither tracking the path of magma migration nor it defines the boundaries of magma storage volumes such as a midcrustal sill. Our conceptual model considers pulsatory magma injection from the upper mantle and its propagation along the Moho. We suggest, within this framework, that the spatial and temporal distributions of earthquake hypocenters reflect hydraulic fracturing processes associated with stress propagation due to magma movement.

Discriminative Feature Selection for Automatic Classification of Volcano-Seismic Signals

Feature extraction is a critical element in automatic pattern classification. In this letter, we propose different sets of parameters for classification of volcano-seismic signals, and the discriminative feature selection (DFS) method is applied for selecting the minimum number of features containing most of the discriminative information. We have applied DFS to a conventional cepstral-based parameterization (with 39 features) and to an extended set of parameters (including 84 features). Classification experiments using seismograms recorded at Colima Volcano (Mexico) show that, for the most complex classifier and using the cepstral-based parameterization, DFS provided a reduction of the error rate from 24.3% (using 39 features) to 15.5% (ten components). When DFS is applied to the extended parameterization, the error rate decreased from 27.9% (84 features) to 13.8% (14 features). These results show the utility of DFS for identifying the best components from the original feature vector and for exploring new parameterizations for the classification of volcano-seismic signals.

An Automatic P-Phase Picking Algorithm Based on Adaptive Multiband Processing

This letter presents a novel picking algorithm which allows an automated determination of the P-phase onset time. The algorithm includes an adaptive multiband processing and noise-reduction techniques to allow a confident onset time estimation in signals strongly affected by background and/or nonstationary noise processes. Results using a set of 3780 computer-generated earthquake-like signals show that the accuracy is much better than that achieved by conventional STA/LTA algorithm. In addition, the accuracy of the proposed method is improved when it is combined with an autoregressive method. An application of the algorithm to a set of 400 natural earthquakes confirms that the combination of both algorithms provides a precise P-phase onset time estimation in real environments, overcoming the limitations associated with the autoregressive method.

A Comparative Study of Dimensionality Reduction Algorithms Applied to Volcano-Seismic Signals

Detection and classification of the different seismic events are important tasks in volcanological observatories. Trying to make these an automatic process is fundamental for the volcanological community. It is crucial to choose how the seismic signal is represented in terms of parameters or features useful for dealing with the automatic classification problem, since the number and type of parameters could be really large leading to the curse of dimensionality issue. Machine learning theory establishes that in order to build a classifier from a labeled database, there should be a compromise between the complexity of the classifier and the size of the database. Since generating a manually labeled database is a tedious work performed by specialists in volcanology, the size of the databases limits the complexity of the classification systems built by them. On the other hand, if the databases could be represented by a reduced, but relevant, number of features, the complexity of the classifier would be simplified. In order to study the problem just described, this paper performs a comparative study of different classical techniques of dimensionality reduction (DR) of the feature set. The algorithms implemented include feature selection techniques as wrappers and filters and methods which directly transform the original feature space into another with lower dimension. All algorithms have been tested using an automatic classification system of volcano-seismic events. The best results have been obtained with the discriminative feature selection (DFS) algorithm which belongs to the set of wrapper methods.

APASVO: A free software tool for automatic P-phase picking and event detection in seismic traces

Abstract The accurate estimation of the arrival time of seismic waves or picking is a problem of major interest in seismic research given its relevance in many seismological applications, such as earthquake source location and active seismic tomography. In the last decades, several automatic picking methods have been proposed with the ultimate goal of implementing picking algorithms whose results are comparable to those obtained by manual picking. In order to facilitate the use of these automated methods in the analysis of seismic traces, this paper presents a new free, open source, software graphical tool, named APASVO, which allows picking tasks in an easy and user-friendly way. The tool also provides event detection functionality, where a relatively imprecise estimation of the onset time is sufficient. The application implements the STA-LTA detection algorithm and the AMPA picking algorithm. An autoregressive AIC-based picking method can also be applied. Besides, this graphical tool is complemented with two additional command line tools, an event picking tool and a synthetic earthquake generator. APASVO is a multiplatform tool that works on Windows, Linux and OS X. The application can process data in a large variety of file formats. It is implemented in Python and relies on well-known scientific computing packages such as ObsPy, NumPy, SciPy and Matplotlib.

Histogram Equalization for Robust Speech Recognition

Optimal Automatic Speech Recognition takes place when the evaluation is done under circumstances identical to those in which the recognition system was trained. In the speech applications demanded in the actual real world this will almost never happen. There are several variability sources which produce mismatches between the training and test conditions. Depending on his physical or emotional state, a speaker will produce sounds with unwanted variations transmitting no acoustic relevant information. The phonetic context of the sounds produced will also introduce undesired variations. Inter-speaker variations must be added to those intra-speaker variations. They are related to the peculiarities of speakers’ vocal track, his gender, his socio-linguistic environment, etc. A third source of variability is constituted by the changes produced in the speaker’s environment and the characteristics of the channel used to communicate. The strategies used to eliminate the group of environmental sources of variation are called

Evaluating robustness of a HMM-based classification system of volcano-seismic events at colima and popocatepetl volcanoes

This work presents a continuous volcano-seismic classification system based in the Hidden Markov Models as solution to recently strong needs for automatic event detection and recognition methods in early warning and monitoring scenarios. Furthermore, our system includes a reliable method to assign confidence measures to the recognized signals in order to evaluate the robustness of the results. Data from the two most active volcanoes have been used to probe the system reliability on a complex joint corpus achieving a recognition accuracy higher than 78% in blind recognition tests.