Packet loss due to misrouted or delayed packets in voice over IP leads to huge voice quality degradation. This paper presents a packet loss concealement algorithm which is inde pendant from the vocoder. This method relies on hidden Marko v model (HMM). A new voicing parameter is introduced to get ove r voiced/unvoiced sound separation and use a unique HMM. Since best parameter for prediction are not necessarily the best ones for synthesis, we introduce two separate vectors: the first one dedicated to the analysis of the signal and the second one featured for the synthesis of missing part. Performances of the proposed system are evaluated on parts of well-known speech corpora, leading to promising results.

This paper studies a new unmixing algorithm for hyperspectral images. Each pixel of the image is modeled as a linear combination of endmembers which are supposed to be random in order to model uncertainties regarding their knowledge. More precisely, endmembers are modeled as Gaussian vectors with known means (resulting from an endmember extraction algorithm such as the famous N-FINDR or VCA algorithm). This paper proposes to estimate the mixture coefficients (referred to as abundances) using a Bayesian algorithm. Suitable priors are assigned to the abundances in order to satisfy positivity and additivity constraints whereas a conjugate prior is chosen for the variance. The computational complexity of the resulting Bayesian estimators is alleviated by constructing an hybrid Gibbs algorithm to generate abundance and variance samples distributed according to the posterior distribution of the unknown parameters. The associated hyperparameter is also generated. The performance of the proposed methodology is evaluated thanks to simulation results conducted on synthetic and real images.

Packet loss due to misrouted or delayed packets in voice over IP leads to huge voice quality degradation. Packet loss con- cealment algorithms try to enhance the quality of the speech. This paper presents a new packet loss concealment algorithm which relies on one hidden Markov model. For this purpose, we introduce a continuous observation vector well-suited for silence, voiced and unvoiced sounds. We show that having a global HMM is relevant for this application. The proposed system is evaluated using standard PESQ score in a real- world application.

In automatic speech analysis, voicing articulatory is often defined as a binary decision: voiced or unvoiced. Lin- guists agree that this articulatory should be continuous. In this paper, we present a new approach to compute a con- tinuous voicing indicator of a speech frame. This voic- ing percentage is then evaluated in both a segmentation process and a speech recognition task. Promising results show that this continuous voicing percentage may be used as a reliable voicing indicator.

Detecting the presence of non-stationarity events in a signal is a challenge that is still not taken up. The aim of this paper is to make a contribution to this key issue. We already proposed a non-stationarity detection defined in time-frequency domain in order to control the invariance of the time-frequency statistics. In this paper, in order to be not limited by the time and frequency resolution of a time-frequency approach, we propose another test in frequency domain. In frequency domain, the problem can be cast by taking advantage of the normalized-variance properties of a spectral estimator when analyzing non-stationary signals. This second test will confirm, invalidate or detect new frequency localizations of non-stationarities. Finally, the main contribution of the paper is to propose a stationary index defined so as to merge the information given by these two tests and to allow an alarm to be raised for a high level of non-stationarities. Applications on real-world signals show the pertinence of this new index.

OURSES project, French acronym for Offer of Services Rural Use using Satellite, proposes three telemedicine applications linked to services for elderly people. It focuses particularly on the use of satellites as a complement to terrestrial technologies to ensure the deployment of tele-services in areas where telecommunication infrastructure is lacking. This paper describes the three telemedicine applications which can be viewed as tele-monitoring systems for elderly people.

The pseudo maximum likelihood estimator allows one to estimate the unknown parameters of Brown’s model for altimeter wave- forms. However, the optimality of this estimator, for instance in terms of minimizing the mean square errors of the unknown para- meters is not guarantied. Thus it is not clear whether there is some space for developing new estimators for the unknown parameters of altimetric signals. This paper derives the Crame ́r-Rao lower bounds of the parameters associated to Brown’s model. These bounds provide the minimum variances of any unbiased estima- tor of these parameters, i.e. a reference in terms of estimation er- ror. A comparison between the mean square errors of the standard estimators and Crame ́r-Rao bounds allows one to evaluate the po- tential gain (in terms of estimation variance) that could be achieved with new estimation strategies.

Several telemedicine applications are proposed within the frame of OURSES project, French acronym for Offer of Rural Use of Services by Satellite, providing services for elderly people. The main objective of this project is to show the interest of using satellites as a complement to terrestrial technologies, in areas where telecommunication infrastructure is lacking or incomplete. This paper describes one of these applications: an ECG monitoring system. This telemonitoring system allows, thanks to a wireless wearable sensor, to detect possible cardiac problems of elderly people. ECG signals are analyzed through signal processing algorithms and if some abnormal condition is detected, an alarm is raised and sent via satellite to the physician's office. The corresponding physician is able to access at any time the recorded ECG signals, whenever he is willing to, in the presence of an alarm or not. This allows a constant monitoring of the elderly people. Tests realized in a real environment have demonstrated the feasibility and the interest of this application.

This paper addresses the problem of classifying altimetric waveforms backscattered from different kinds of surfaces including oceans, ices, deserts and forests. Appropriate features associated with altimetric radar waveforms are first introduced for this classification. These features are completed by radiometer temperatures and pre-processed using a linear discriminant analysis for dimensionality reduction. The classification of altimetric waveforms is finally achieved using the resulting pre-processed vector with reduced dimension. Different classification strategies are finally considered. These strategies are based on the nearest mean rule, the nearest neighbor method or on the multilayer perceptron. Various simulation results illustrate the performance of the proposed classifier.

This paper studies a Bayesian algorithm for estimating the parameters associated to Brown's model. The joint posterior distribution of the unknown parameter vector (amplitude, epoch and significant wave height) associated with this model is derived. This posterior is too complex to obtain closed form expressions of the minimum mean square error and the maximum a posteriori estimators. We propose to sample according to this distribution using an hybrid Metropolis within Gibbs algorithm. The simulated samples are then used to estimate the unknown parameters of Brown's model. The proposed strategy provides better estimations than the standard maximum likelihood estimator at the price of an increased computational cost.