Search

This paper presents our approach for automatic speech recognition (ASR) of overlapping speech. Our system consists of two principal components: a speech separation component and a feature estmation component. In the speech separation phase, we first estimated the speaker΄s position, and then the speaker location information is used in a GSC-configured beamformer with a minimum mutual information (MMI) criterion, followed by a Zelinski and binary-masking post-filter, to separate the speech of different speakers. In the feature estimation phase, the neural networks are trained to learn the mapping from the features extracted from the pre-separated speech to those extracted from the close-talking microphone speech signal. The outputs of the neural networks are then used to generate acoustic features, which are subsequently used in acoustic model adaptation and system evaluation. The proposed approach is evaluated through ASR experiments on the PASCAL Speech Separation Challenge II (SSC2) corpus. We demonstrate that our system provides large improvements in recognition accuracy compared with a single distant microphone case and the performance of ASR system can be significantly improved both through the use of MMI beamforming and feature mapping approaches. [ABSTRACT FROM AUTHOR]

The PASCAL Speech Separation Challenge (SSC) is based on a corpus of sentences from the Wall Street Journal task read by two speakers simultaneously and captured with two circular eight-channel microphone arrays. This work describes our system for the recognition of such simultaneous speech. Our system has four principal components: A person tracker returns the locations of both active speakers, as well as segmentation information for each utterance, which are often of unequal length; two beamformers in generalized sidelobe canceller (GSC) configuration separate the simultaneous speech by setting their active weight vectors according to a minimum mutual information (MMI) criterion; a postfilter and binary mask operating on the outputs of the beamformers further enhance the separated speech; and finally an automatic speech recognition (ASR) engine based on a weighted finite-state transducer (WFST) returns the most likely word hypotheses for the separated streams. In addition to optimizing each of these components, we investigated the effect of the filter bank design used to perform subband analysis and synthesis during beamforming. On the SSC development data, our system achieved a word error rate of 39.6%. [ABSTRACT FROM AUTHOR]

The high dimensionality of functional magnetic resonance imaging (fMRI) data presents major challenges to fMRI pattern classification. Directly applying standard classifiers often results in overfitting, which limits the generalizability of the results. In this paper, we propose a new group of classifiers, ˵Generalized Sparse Classifiers″ (GSC), to alleviate this overfitting problem. GSC draws upon the recognition that numerous standard classifiers can be reformulated under a regression framework, which enables state-of-the-art regularization techniques, e.g. elastic net, to be directly employed. Building on this regularized regression framework, we exploit an extension of elastic net that permits general properties, such as spatial smoothness, to be integrated. GSC thus facilitates simultaneous sparse feature selection and classification, while providing greater flexibility in the choice of penalties. We validate on real fMRI data and demonstrate how explicitly modeling spatial correlations inherent in brain activity using GSC can provide superior predictive performance and interpretability over standard classifiers. [ABSTRACT FROM AUTHOR]

As far as we know there is not a definition of dependency in a formal setting: to fill this gap we propose in this paper a state based formalism called (network of) Dependent Automata, that consider dependencies as central elements. When used for modelling interdependencies in critical infrastructures, each infrastructure is modelled as a Dependent Automaton, that accounts for local behaviour and for dependencies from and to other infrastructures, while the whole system is obtained by composition of the automata of the infrastructures considered. [ABSTRACT FROM AUTHOR]

As seen throughout the text, the major functionality of a microphone-array system is to reduce noise, thereby enhancing a desired information-bearing speech signal. The term noise, in general, refers to any unwanted signal that interferes with measurement, processing, and communication of the desired speech signal. This broad-sense definition of noise, however, is too encompassing as it masks many important technical aspects of the real problem. To enable better modeling and removal of the effects of noise in the context of microphone array processing, it is advantageous to break the general definition into the following three subcategories: additive noise originating from various ambient sound sources, interfering signals from concurrent competing sources, and reverberation caused by multipath propagation introduced by an enclosure. We have seen from the previous chapters that the use of a microphone array together with proper beamforming techniques can reduce the effect of additive noise. This chapter continues to explore beamforming techniques, with a focus on interference suppression and speech dereverberation. Different from the traditional way of treating beamforming as purely spatial filtering, this chapter studies the problem from a more physically meaningful multiple-input multiple-output (MIMO) signal processing perspective. A general framework based on the MIMO channel impulse responses will be developed. Under this framework, we study different algorithms including their underlying principles and intrinsic connections. [ABSTRACT FROM AUTHOR]

In this research we investigate the relationship between emotion and cooperation in dialogue tasks. It is an area were still many unsolved questions are present. One of the main open issues is the labeling of ˵blended″ emotions and their recognition. Usually there is a low agreement among raters in labeling and naming emotions and surprisingly emotion recognition is higher in a condition of modality deprivation (only acoustic or only visual vs. bimodal). Because of this previous results we don΄t ask raters to directly label emotions, but to use a small set of features (as lips or eyebrows shape) to annotate our corpus. The analyzed materials come from an audiovisual corpus of Map Task dialogues elicited with a script. We point out the ˵emotive″ tokens by simultaneous recordings of the phsychophysiological indexes (ElectroCardioGram ECG, Galvanic Skin Conductance GSC, ElectroMyoGraphy EMG). After this selection we annotate each token with our multimodal annotation scheme. Each annotation will lead to a cluster of signals identifying the emotion corresponding to a cooperative/non cooperative level; the last step involves agreement among coders and reliability of the emotion description. Future research will deal with brain imaging experiment on the effect of putting emotions into words and the role of context in emotion recognition. [ABSTRACT FROM AUTHOR]

Delayed onset muscle soreness (DOMS) is a common experience following unaccustomed eccentric exercise. DOMS and associated force deficits may limit optimal performance in subsequent days. The cause of DOMS remains poorly understood, thus there is no effective treatment. Graduated compression stockings (GCS) are a commonly used intervention believed to diminish DOMS. The purpose of this study was to determine if GCS after eccentric walking exercise minimizes DOMS and associated deficits (e.g. muscle force capacity). Eight healthy subjects (age 26±4 yrs, height 175±8 cm, weight 70±5 kg) volunteered to perform a single bout of backward downhill walking exercise (duration 30 min, velocity 1 m.s-1, negative grade-25%, load 12% of body weight). Following walking exercise, subjects were required to wear 5 hours per day for 3 consecutive days GSC (SupportivTM) on one leg while the second was used as control. Muscle soreness and neuromuscular measures (M-wave, peak twitch, maximal voluntary torque or MVT) were taken pre and postwalk, then 2, 24, 48 and 72 hours post-walking exercise for the two legs. There was a 28% reduction in DOMS 72 h after exercise when wearing GCS (P<0.05) than in the control leg. Immediately after exercise there was a 15% decrease in MVT of the plantar flexors in both legs partly attributable to an alteration in contractile properties (-22% in electrically evoked mechanical twitch). In leg wearing GCS, MVT starts to recover while the contractile properties had significantly recovered within 24 h but not in the control leg. In the current study, GCS might have had the effect of compressing the muscle tissue to such an extent that less structural damage occurred relative to a control condition. GCS accelerated the recovery of the muscle force capacity at 24 hours beyond that achieved by the control condition. [ABSTRACT FROM AUTHOR]

Copyright of Erfolgreiche Umsetzung von Work-life Balance in Organisationen is the property of Springer Nature / Books and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

This paper presents a combined handsfree speech enhancement method based on a spatialpost-filter. The scheme uses a linear microphone array to capture a speech signal that has been corrupted by babble noise, car noise, and interference signals. Simulation results for real environment show that the proposed structure achieves a maximum interference suppression of 12 dB, an improvement of 6 dB over the delay and sum beamformer. Furthermore, the system is robust in the presence of distortion as opposed to the generalized sidelobe canceller. The subjective evaluation has shown that the combined system of delay and sum with the minimum mean square error estimator using a noncausal signal to noise ratio (SNR) estimator obtained 3.8 points on a fivepoint. [ABSTRACT FROM AUTHOR]

The ubiquitous noise reduction / speech enhancement problem has gained an increasing interest in recent years. This is due both to progress made by microphone-array systems and to the successful introduction of perceptual models. In the last decade, several methods incorporating psychoacoustic criteria in single channel speech enhancement systems have been proposed, however very few works exploit these features in the multichannel case. In this paper we present a novel psychoacoustically motivated, multichannel speech enhancement system that exploits spatial information and psychoacoustic concepts. The proposed framework offers enhanced flexibility allowing for a multitude of perceptually-based post-filtering solutions. Moreover, the system has been devised on a frame-by-frame basis to facilitate real-time implementation. Objective performance measures and informal subjective listening tests for the case of speech signals corrupted with real car and F-16 cockpit noise demonstrate enhanced performance of the proposed speech enhancement system in terms of musical residual noise reduction compared to conventional multichannel techniques. [ABSTRACT FROM AUTHOR]

The formations of the blue straggler stars and the FK Com-type stars are unsolved problems in stellar astrophysics. One of the possibilities for their formations is from the coalescence of W UMa-type overcontact binary systems. Therefore, deep (f > 50%), low-mass ratio (q < 0.25) overcontact binary stars are a very important source to understand the phenomena of Blue Straggler/FK Com-type stars. Recently, 12 W UMa-type binary stars, FG Hya, GR Vir, IK Per, TV Mus, CU Tau, V857 Her, V410 Aur, XY Boo, SX CrV, QX And, GSC 619-232, and AH Cnc, were investigated photometrically. Apart fromTV Mus, XY boo, and GSC 619-232, new observations of the other 9 binaries were obtained. Complete light curves of the 10 systems, FG Hya, GR Vir, IK Per, TV Mus, CU Tau, V857 Her, GSC 619-232, V410 Aur, XY Boo, and AH Cnc, were analyzed with the 2003 version of the W-D code. It is shown that all of those systems are deep (f > 50%), low-mass ratio (q < 0.25) overcontact binary stars. We found that the system GSC 619-232 has the highest degree of overcontact (f = 93.4%). The derived photometric mass ratio of V857 Her, q = 0.0653, indicates that it is the lowest-mass ratio system among W UMa-type binaries. Of the 12 sample stars, long-term period changes of 11 systems were found. About 58% (seven) of the sample binaries show cyclic period oscillation. No cyclic period changes were discovered for the other 5 systems, which may be caused by the short observational time interval or by insufficient observations. Therefore, we think that all W UMa-type binary stars may contain cyclic period variations. By considering the long-term period changes (both increase and decrease) of those binary stars, we proposed two evolutionary scenarios evolving from deep, low-mass ratio overcontact binaries into Blue Straggler/FK Com-type stars. [ABSTRACT FROM AUTHOR]

Multi-channel noise reduction has been widely researched to reduce acoustic noise signals and to improve the performance of many speech applications in noisy environments. In this paper, we first introduce the state-of-the-art multi-channel noise reduction methods, especially beamforming based methods, and discuss their performance limitations. Subsequently, we present a multi-channel noise reduction system we are developing that consists of localized noise suppression by microphone array and non-localized noise suppression by post-filtering. Experimental results are also presented to show the benefits of our developed noise reduction system with respect to the traditional algorithms in terms of speech recognition rate. Some suggestions are finally presented for the further research. [ABSTRACT FROM AUTHOR]

In many speech communication applications a microphone array is available nowadays, such that multi-microphone speech enhancement techniques can be used instead of single-microphone speech enhancement techniques. A wellknown multi-microphone speech enhancement technique is the generalized sidelobe canceller (GSC), which is however quite sensitive to signal model errors, such as microphone mismatch. This chapter discusses a more robust technique called the spatially pre-processed speech distortion weighted multichannel Wiener filter (SPSDW-MWF), which takes speech distortion due to signal model errors explicitly into account in its design criterion, and which encompasses the standard GSC as a special case. In addition, a novel frequency-domain criterion for the SDW-MWF is presented, from which several — existing and novel — adaptive frequency-domain algorithms can be derived for implementing the SDW-MWF. The noise reduction performance and the robustness of these adaptive algorithms is investigated for a hearing aid application. Using experimental results with a small-sized microphone array, it is shown that the SP-SDW-MWF is more robust against signal model errors than the GSC, both in stationary and in changing noise scenarios. [ABSTRACT FROM AUTHOR]