A multimodal alignment framework for spoken documents

We present a multimodal document alignment framework, which highlights existing alignment relationships between documents that are discussed and recorded during multimedia events such as meetings. These relationships that should help indexing the archives of these events are detected using various techniques from natural language processing and information retrieval. The main alignment strategies studied are based on thematic, quotation and reference relationships. At the analysis level, the alignment framework was applied at several levels of granularity of documents, requiring specific document segmentation techniques. Our framework that is language independent was evaluated on corpora in French and English, including meetings and scientific presentations. The satisfactory evaluation results obtained at several stages show the importance of our approach in bridging the gap between meeting documents, independently from the language and domain. They highlight also the utility of the multimodal alignment in advanced applications, e.g. multimedia document browsing, content-based / temporal-based searching, etc.     

Facilitating the migration to the microservice architecture via model-driven reverse engineering and reinforcement learning

The microservice architecture has gained remarkable attention in recent years. Microservices allow developers to implement and deploy independent services, so they are a naturally effective architecture for continuously deployed systems. Because of this, several organizations are undertaking the costly process of manually migrating their traditional software architectures to microservices. The research in this paper aims at facilitating the migration from monolithic software architectures to microservices. We propose a framework which enables software developers/architects to migrate their software systems more efficiently by helping them remodularize the source code of their systems. The framework leverages model-driven reverse engineering to obtain a model of the legacy system and reinforcement learning to propose a mapping of this model toward a set of microservices.

     

Insight of EDX analysis and EFTEM: Are spherocrystals located in strombidae digestive gland implied in detoxification of trace metals?

Digestive tubules of Strombidae are composed by three cell types: digestive cells, vacuolated cells, and crypt cells. The last one is characterized by the presence of intracellular granules identified as spherocrystals. Such structures are known to occur in basophilic cells of gastropod digestive gland, where they are supposed to be involved in the regulation of some minerals and in detoxification. In this study, energy‐dispersive X‐ray analysis (EDX) and energy filtered transmission electron microscopy (EFTEM) were used to determine the elemental content of spherocrystals in two Strombidae, Strombus gigas and Strombus pugilis. In freshly collected individuals of both species, the following elements were detected: Ca, Fe, Mg, P, and Zn. Aluminum and Mn were also detected in S. gigas. Their presence in spherocrystals indicates that, in Strombidae, spherocrystals are involved in the regulation of minerals and essential trace metals. In order to answer the question “are spherocrystals involved in nonessential trace metals scavenging?,” artificial cadmium and lead exposure by both waterborne and dietary pathways was applied to S. pugilis. No evidence of cadmium (Cd(NO₃)₂) or lead (Pb(NO₃)₂) provided by food was found in spherocrystals. Cadmium provided in water (Cd(NO₃)₂ and CdCl₂) causes structural modifications of the digestive gland; however, this element was not trapped in spherocrystals. These results suggest that spherocrystals are not involved in detoxification of such nonessential trace metals. Microsc. Res. Tech., 2011. © 2011 Wiley‐Liss, Inc.     

Assessment of the Wind Field in the East Coast Algerian Regions for the Installation of Wind Farms

The available electricity generated by a wind power generation system depends on mean wind speed, standard deviation of wind speed and the location of installation. Since year-to-year variation on annual mean wind speed is hard to predict, wind speed variations during a year can be well characterized in terms of a probability distribution function, as well Weibull distribution has been one of the most commonly used, accepted and recommended distribution to determine wind energy potential. In this study, the two Weibull parameters of the wind speed distribution function （the shape parameter k （dimensionless） and the scale parameter c （m/s））, were computed from the wind speed data for Algerian east coastal regions, recording over a 1 l-year period （1995-2005）. It was found that the numerical values of both Weibull parameters （k and c） vary over a wide range. The yearly values ofk range from 1.20 to 1.94, while those of c are in the range from 4.31 to 1.94. To minimize the uncertainties of statistical calculation, a spatial representation indicating distribution and occurrence frequency the direction from which the wind comes, appears a very primordial step. Over the whole valid data during the study period, the compass shows that there is no dominant direction marked. However, we can identify a preferred wind direction. The statistical results correspond to the analysis of the rose compass.     

Robust regression fusion of GMM-UBM and GMM-SVM normalized scores using G729 bit-stream for speaker recognition over IP

A novel approach, based on robust regression with normalized score fusion (namely Normalized Scores following Robust Regression Fusion: NSRRF), is proposed for enhancement of speaker recognition over IP networks, which can be used both in Network Speaker Recognition (NSR) and Distributed Speaker Recognition (DSR) systems. In this framework, it is basically assumed that the speech must be encoded by G729 coder in client side, and then, transmitted at a server side, where the ASR systems are located. The Universal Background Gaussian Mixture Model (GMM-UBM) and Gaussian Supervector (GMM-SVM) with normalized scores are used for speaker recognition. In this work, Mel Frequency Cepstral Coefficient (MFCC) and Linear Prediction Cepstral Coefficient (LPCC), both of these features are derived from Line Spectral Pairs (LSP) extracted from G729 bit-stream over IP, constitute the features vectors. Experimental results, conducted with the LIA SpkDet system based on the ALIZE platform3 using ARADIGITS database, have shown in first that the proposed method using features extracted directly from G729 bit-stream reduces significantly the error rate and outperforms the baseline system in ASR over IP based on the resynthesized (reconstructed) speech obtained from the G729 decoder. In addition, the obtained results show that the proposed approach, based on scores normalization following robust regression fusion technique, achieves the best result and outperform the conventional ASR over IP network.     

A modified particle swarm optimisation algorithm to solve the part feeding problem at assembly lines

The Assembly Line Part Feeding Problem (ALPFP) is a complex combinatorial optimisation problem concerned with the delivery of the required parts to the assembly workstations in the right quantities at the right time. Solving the ALPFP includes simultaneously solving two sub-problems, namely tour scheduling and tow-train loading. In this article, we first define the problem and formulate it as a multi-objective mixed-integer linear programming model. Then, we carry out a complexity analysis, proving the ALPFP to be NP-complete. A modified particle swarm optimisation (MPSO) algorithm incorporating mutation as part of the position updating scheme is subsequently proposed. The MPSO is capable of finding very good solutions with small time requirements. Computational results are reported, demonstrating the efficiency and effectiveness of the proposed MPSO.     

Markovian modelling and Fisher distribution for unsupervised classification of radar images

Statistical segmentation techniques based on hidden Markov field modelling have generated considerable interest in past years. They take contextual information into account in a particularly elegant and rigorous way. Although these models have been thoroughly tested, they can fail in some cases such as the non-stationary one. In this article, we propose use of the recently developed triplet Markov field, which models non-stationary images, and that of Fisher distribution, which is adapted to a wide range of surfaces for modelling synthetic aperture radar (SAR) image noise. Examples illustrate the difference between the approach proposed and classical ones. Various experiments indicate that the new model and its associated unsupervised algorithm perform better than classical ones.