A new approach of speaker clustering based on the stereophonic differential energy

In this paper, we present a new approach of speech clustering with regards of the speaker identity. It consists in grouping the homogeneous speech segments that are obtained at the end of the segmentation process, by using the spatial information provided by the stereophonic speech signals. The proposed method uses the differential energy of the two stereophonic signals collected by two cardioid microphones, in order to cluster all the speech segments that belong to the same speaker. The total number of clusters obtained at the end should be equal to the real number of speakers present in the meeting room and each cluster should contain the global intervention of only one speaker. The proposed system is suitable for debates or multi-conferences for which the speakers are located at fixed positions. Basically, our approach tries to make a speaker localization with regards to the position of the microphones, taken as a spatial reference. Based on this localization, the new proposed method can recognize the speaker identity of any speech segment during the meeting. So, the intervention of each speaker is automatically detected and assigned to him by estimating his relative position. In a purpose of comparison, two types of clustering methods have been implemented and experimented: the new approach, which we called Energy Differential based Spatial Clustering (EDSC) and a classical statistical approach called “Mono-Gaussian based Sequential Clustering” (MGSC). Experiments of speaker clustering are done on a stereophonic speech corpus called DB15, composed of 15 stereophonic scenarios of about 3.5 minutes each. Every scenario corresponds to a free discussion between two or three speakers seated at fixed positions in the meeting room. Results show the outstanding performances of the new approach in terms of precision and speed, especially for short speech segments, where most of clustering techniques present a strong failure.     

A genetic local tuning algorithm for a class of combinatorialnetworks design problems

Experimental evidences of many genetic algorithm (GA) researchers is that hybridizing a GA with a local search (LS) heuristic is beneficial. It combines the ability of the GA to widely sample a search space with a local search hill-climbing ability. This letter presents a genetic local search (GALS) mechanism applied on two stages on the initial genetic population. An elite nondominated set of solutions is selected, an intermediate population (IP) composed of the elite and the improved solutions by natural genetic operators is constructed and then a Nelder and Mead (1965) simplex downhill method (SDM) is applied to some solutions of the IP. Experimental results from solving a 20-nodes topology design and capacity assignment (TDCA) problem suggest that our approach provides superior results compared to four simple GA implementations found in the literature     

Designing communication network topologies using steady-stategenetic algorithms

This article presents the application of steady state genetic algorithms (SSGA) to minimize the total installation cost of a communication network by optimally designing the topology layout and assigning the corresponding capacities (TDCA problem). This highly constrained optimization problem is shown to be better solved using GAs. A binary representation of links between node pairs is developed and tested on a network of 20 nodes. Improved results, both in terms of network cost, performance and computation speed, are obtained when comparing with existing heuristic approaches     

A pertinent learning machine input feature for speaker discrimination by voice

This research work is a part of a global project of speech indexing entitled ISDS and concerns more particularly two machine learning classifier types: Neural Networks (NN) and Support Vector Machines (SVM), which are used by that project. However, in the present paper, we will only deal with the problem of speaker discrimination using a new relative reduced modelization for the speaker, restricting then our analysis to the new relative speaker characteristic used as input feature of the learning machines (NN and SVM). Speaker discrimination consists in checking whether two speech signals belong to the same speaker or not, by using some features of the speaker directly from his own speech. Our new proposed feature is based on a relative characterization of the speaker, called Relative Speaker Characteristic (RSC) and is well adapted for NN and SVM trainings. RSC consists in modeling one speaker relatively to another one, meaning that each speaker model is determined from both its speech signal and its dual speech. This investigation shows that the relative model, used as input of the classifier, optimizes the training, by speeding up the learning time and enhancing the discrimination accuracy of that classifier.     

Speaker discrimination based on fuzzy fusion and feature reduction techniques

In this paper, we propose a research work on speaker discrimination using a multi-classifier fusion with focus on feature reduction effects. Speaker discrimination consists in the automatic distinction between two speakers using the vocal characteristics of their speeches. A number of features are extracted using Mel Frequency Spectral Coefficients and then reduced using Relative Speaker Characteristic (RSC) along with the Principal Components Analysis (PCA). Several classification methods are implemented to ensure the discrimination task. Since different classifiers are employed, two fusion algorithms at the decision level, referred to as Weighted Fusion and Fuzzy Fusion, are proposed to boost the classification performances. These algorithms are based on the weighting of the different classifiers outputs. Furthermore, the effects of speaker gender and feature reduction on the speaker discrimination task have been examined too. The evaluation of our approaches was conducted on a subset of Hub-4 Broadcast-News. The experimental results have shown that the speaker discrimination accuracy is improved by 5–15% using the (RSC–PCA) feature reduction. In addition, the proposed fusion methods recorded an improvement of about 10% compared to the individual scores of the classifiers. Finally, we noticed that the gender has an important impact on the discrimination performances.     

A novel robust Arabic light stemmer

Abstract

The stemming is the process of transforming a word into its root or stem, hence, it is considered as a crucial pre-processing step before tackling any task of natural language processing or information retrieval. However, in the case of Arabic language, finding an effective stemming algorithm seems to be quite difficult, since the Arabic language has a specific morphology, which is different from many other languages. Although, there exist several algorithms in literature addressing the Arabic stemming issue, unfortunately, most of them are restricted to a limited number of words, present some confusions between original letters and affixes, and usually employ dictionary of words or patterns. For that purpose, we propose the design and implementation of a novel Arabic light stemmer, which is based on some new rules for stripping prefixes, suffixes and infixes in a smart way. And in our knowledge, it is the first work dealing with Arabic infixes with regards to their irregular rules. The empirical evaluation was conducted on a new Arabic data-set (called ARASTEM), which was conceived and collected from several Arabic discussion forums containing dialectical Arabic and modern pseudo-Arabic languages. Hence, we present a comparative investigation between our new stemmer and other existing stemmers using Paice’s parameters, namely: Under Stemming Index (UI), Over Stemming Index (OI) and Stemming Weight (SW). Results show that the proposed Arabic light stemmer maintains consistently high performances and outperforms several existing light stemmers.