Dissertations / Theses on the topic 'Reconnaissance de la langue'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Reconnaissance de la langue.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
Nguyen, Quoc Cuong. "Reconnaissance de la parole en langue vietnamienne." Grenoble INPG, 2002. http://www.theses.fr/2002INPG0044.
Burger, Thomas. "Reconnaissance automatique des gestes de la langue française parlée complétée." Phd thesis, Grenoble INPG, 2007. http://tel.archives-ouvertes.fr/tel-00203360.
Duroselle, Raphaël. "Robustesse au canal des systèmes de reconnaissance de la langue." Electronic Thesis or Diss., Université de Lorraine, 2021. http://www.theses.fr/2021LORR0250.
Language recognition is the task of predicting the language used in a test speech utterance. Since 2017, the best performing systems have been based on a deep neural network which is trained to predict language labels for the whole utterance. These systems suffer from a drop in performance when they are exposed to a change of the transmission channel between train and test data. The goal of this thesis is to investigate approaches to limit this performance drop, for these new systems.An increase in the invariance, with respect to the transmission channel, of the representations used by the neural network can increase the robustness of the system. We show that the regularization of the loss function used to train the neural network is an efficient approach to increase invariance. Two kinds of regularization functions are analysed. Divergence measures between domains reduce effectively the variability between known domains, they can also be used to incorporate unlabeled data into the training set in a semi-supervised learning framework. Metric learning cost functions are able to reduce unknown variabilities within the training set. We show how this regularization method can be enforced for three practical learning settings: unsupervised domain adaptation, multi-domain learning and domain generalization.During this work, we have designed methods for analyzing the quality of the representations. They aim at evaluating the variability of the representations induced by the transmission channel and to compare it to the variability that caused the language. Two tools are proposed: ratio between inter class and intra class covariance matrices and divergence measures between groups of representations. With these tools, we quantitatively evaluate the robustness to a change of transmission channel of the representations and analyse the effect of the regularization functions over the space of representations. We understand that an increase in invariance between channels can lead to more discriminative representations between languages and consequently to an increase in performance over each transmission channel.Finally, we contribute to the improvement of the training recipe of another module of the system, the bottleneck feature extractor. We replace it with a multilingual end-to-end automatic speech recognition neural network. It achieves a simiar performance as a traditional bottleneck feature extractor with a simplified training recipe. The use of data augmentation and regularization methods improves further this module. Moreover we show that a performance gain can be achieved with a joint training of the bottleneck feature extractor along with the language identification neural network. This paves the way to the application of the proposed regularization loss functions to the two modules jointly
BRAFFORT, ANNE. "Reconnaissance et comprehension de gestes, application a la langue des signes." Paris 11, 1996. http://www.theses.fr/1996PA112229.
Ben, Mlouka Monia. "Le référencement en langue des signes : analyse et reconnaissance du pointé." Toulouse 3, 2014. http://thesesups.ups-tlse.fr/2676/.
This thesis focuses on the role and analysis of gaze in sign language where it plays an important role. In any language, the gaze keeps the communication relationship. In addition to that, it allows structuring a sign language discourse or interaction between signers, by investing in complex linguistic features. We focus on the role of reference, which is to put the focus on an element of the discourse. In sign language, the components of the discourse are localized in the signing space; thus putting the focus on an element of discourse which is to identify and activate its spatial location (locus), which will mobilize one or more body parts, hands, shoulders, head and eyes. We therefore analyzed the concept of reference in its manual and / or non- manual gestures and set up a reference-based recognition system that takes as input a video in sign language. The recognition system consists of three steps: - 3D modeling of the concept of reference. - The transformation of the 3D model into a 2D model useable by a 2D recognition system. - The detection system, which uses this 2D model. Modeling involves the extraction of gestural characteristics of the concept of reference from corpus consisted of 3D motion capture and gaze and manually annotated videos and the temporal pattern of time lags between motions. Modeling concerns the description of body parts that play a role in reference and the quantification of their gestural. The resulting models describe: 1) The dynamic movement of the dominant hand and 2) the distances between body parts and locus and 3) the time lags between the beginning of motions. The implementation of the recognition method integrates these 3D models. Since the resulting models are three-dimensional and the recognition system has, as input, a 2D video, we propose a transformation of 3D models to 2D to allow their use in the analysis of 2D video and in pattern recognition of reference structures. We can then apply a recognition algorithm to the 2D video corpus. The recognition results are a set of time slots with two main variants of reference. This pioneering work on the characterization and detection of references structures would need to be applied on much larger corpus, consistent and rich and more sophisticated classification methods. However, it allowed to make a reusable methodology of analysis
Darcy, Isabelle. "Assimilation phonologique et reconnaissance des mots." Paris, EHESS, 2003. http://www.theses.fr/2003EHES0131.
Phonological assimilation may change the acoustic shape of words according to the influence of surrounding segments. Word recognition has to cope with this systematic variability which characterizes spoken language. Three word-detection experiments in French, English and American English show that listeners use implicit phonological knowledge of the assimilation processes that exist in their language. Moreover, this knowledge of the assimilation independently of the activation of a lexical form, as the same compensation patterns have been observed in words and in non-words. This compensation mechanism shows some flexibility : comparing French and American second language (L2) learners shows that beginners use the phonological system of their native language to compensate for assimilation processes that exist in L2, but that more advanced learners develop a specific system dedicated to L2 within a few years of exposure
Masmoudi, Dammak Abir. "Approche hybride pour la reconnaissance automatique de la parole en langue arabe." Thesis, Le Mans, 2016. http://www.theses.fr/2016LEMA1040/document.
The development of a speech recognition system requires the availability of a large amount of resources namely, large corpora of text and speech, a dictionary of pronunciation. Nevertheless, these resources are not available directly for Arabic dialects. As a result, the development of a SRAP for Arabic dialects is fraught with many difficulties, namely the lack of large amounts of resources and the absence of a standard spelling as these dialects are spoken and not written. In this perspective, the work of this thesis is part of the development of a SRAP for the Tunisian dialect. A first part of the contributions consists in developing a variant of CODA (Conventional Orthography for Arabic Dialectal) for the Tunisian dialect. In fact, this convention is designed to provide a detailed description of the guidelines applied to the Tunisian dialect. Given the guidelines of CODA, we have created our corpus TARIC: Corpus of the interaction of the railways of the Tunisian Arab in the field of SNCFT. In addition to these resources, the pronunciation dictionary is indispensable for the development of a peech recognition system. In this regard, in the second part of the contributions, we aim at the creation of a system called conversion(Grapheme-Phonème) G2P which allows to automatically generate this phonetic dictionary. All these resources described before are used to adapt a SRAP for the MSA of the LIUM laboratory to the Tunisian dialect in the field of SNCFT. The evaluation of our system gave rise to WER of 22.6% on the test set
SAEB, MOHAMMAD. "Analyse morpho-syntaxique automatique de la langue persane - reconnaissance des syntagmes nominaux." Lyon 2, 1990. http://www.theses.fr/1990LYO20055.
The aim of this study is the eleboration of a linguistic model for the computing of persian. This model is based upon a grammar and a lexicon which can be querried to check the existence of the form in the text submitted to analysis. Based upon this model, the morpho-syntactic analyser generated by the c language in unix operating system, provides the list of all the forms included in a text written in persian. In this model a morphological category is assigend to each form in the text. Noun pharses are retrieved from the above mentioned list. The lexicon is conceived of according to the relational model and implemented upon the informix data base processing system
Pham, Trong-Ton. "MODELE DE GRAPHE ET MODELE DE LANGUE POUR LA RECONNAISSANCE DE SCENES VISUELLES." Phd thesis, Université de Grenoble, 2010. http://tel.archives-ouvertes.fr/tel-00599927.
Bayeh, Rania. "Reconnaissance de la parole multilingue : adaptation de modèles acoustiques vers une langue cible." Paris, Télécom ParisTech, 2009. http://www.theses.fr/2009ENST0060.
Speech processing has become a key technology where different automatic speech recognition (ASR) systems are available for popular languages. With the constant interaction of different cultures, not all users of such systems are native speakers & conversations are often a mixture of several languages which is challenging for ASR. Therefore, a multilingual ASR system is needed. This thesis focuses on efficiently porting the acoustic models (AM) of an under resourced target language using the acoustic models of a more resourced source language with the goal of universal acoustic modeling. Different approaches are suggested & tested for porting models for the recognition of Modern Standard Arabic starting from French for different types of speech & applications. Porting includes the association of speech units, initialization & adaptation of AM. Initially, methods are proposed for the creation of one-to-one phone associations by a human expert or using an automatic data-driven approach. Initialization is done at the context independent level by copying Hidden Markov Models (HMM) target language phone models from a source language HMM phone model based on these associations. Resulting models are adapted using different amounts of target language data. Then, novel methods for one-to-many associations are introduced & multi-path models are used for initialization. Moreover, since the superiority of context dependency extends to cross-lingual & multilingual, different approaches are proposed to create context dependent AM for the under resourced target language using robust AM from a source language. Approaches are also validated for a new language, Colloquial Levantine Arabic
Metzger, Jean-Paul. "Syntagmes nominaux et information textuelle : reconnaissance automatique et représentation." Lyon 1, 1988. http://www.theses.fr/1988LYO19013.
MAHMOUDI, SEYED MOHAMM. "Contribution au traitement automatique de la langue persane : analyse et reconnaissance des syntagmes nominaux." Lyon 2, 1994. http://www.theses.fr/1994LYO20070.
The aim of this thesis is the conception and realisation of a morpho-syntaxic parser of Persian designed for applications to automatic indexing and computer-assisted instruction (or learning) of the language (cai or cal). One of the chief extensions to this research is the automatic processing of natural language by means of artificial intelligence systems. The main interest of this contribution is to study the automatic recognition of noun phrases in Persian. Each stage of the parsing is described in a program in Prolog language (Turbo-Prolog). The whole of the lexical datas necessary for the categorisation of morpho-syntaxic forms is presented as a database
Aboutabit, Noureddine. "Reconnaissance de la Langue Française Parlée Complété (LPC) : décodage phonétique des gestes main-lèvres." Phd thesis, Grenoble INPG, 2007. http://www.theses.fr/2007INPG0153.
Cued Speech (CS) is a visual communication system that uses handshapes placed in different positions near the face, in combination with the natural speech lip-reading, to enhance speech perception from visual input for deaf people. In this system, the speaker moves his hand in close relation with speech. Handshapes are designed to distinguish among consonants whereas hand positions are used to distinguish among vowels. Due to the CS system, both manual and lip flows produced by the CS speaker carry a part of the phonetic information. This work presents at first a method for the automatic coding of the manual flow in term of CS hand positions and CS handshapes. Then the lip-shape classification of the vowels and the consonants is discussed. The labial flow is composed of the temporal variations of lip parameters extracted from the inner and the outer contours of the lips. This work will show how the distribution of lip parameters inside each group of CS hand positions allows vowel discrimination. A classification method based on Gaussian modeling is presented and results demonstrate a good performance of this classification (89% as test score). The vocalic context is taken into account in the case of the consonants, with the use of HMM for the modeling of the lip transition from the consonant towards the vowel (80 % as test scores in term of CV visemes). Finally, the modeling of the lip information and the coding of the manual flow are included in a “Master-Slave” fusion model for recognition of the vowels and the consonants in the CS context. The fusion model integrates the temporal constraints of the CS production and perception. This work is thus also a first contribution to the modeling of the CS system from the perceptive point of view
Aboutabit, Noureddine. "Reconnaissance de la Langue Française Parlée Complété (LPC) : décodage phonétique des gestes main-lèvres." Phd thesis, Grenoble INPG, 2007. http://tel.archives-ouvertes.fr/tel-00270162.
Besse, Anne-Sophie. "Caractéristiques des langues et apprentissage de la lecture en langue première et en français langue seconde : perspective évolutive et comparative entre l'arabe et le portugais." Phd thesis, Université Rennes 2, 2007. http://tel.archives-ouvertes.fr/tel-00267662.
David, Sophie. "Les unités nominales polylexicales : éléments de description et reconnaissance automatique." Paris 7, 1994. http://www.theses.fr/1993PA070108.
Mesfar, Slim. "Analyse morpho-syntaxique automatique et reconnaissance des entités nommées en arabe standard." Besançon, 2008. http://www.theses.fr/2008BESA1022.
The Arabic language, although very important by the number of its speakers, it presents special morpho-syntactic phenomena. This particularity is mainly related to the inflectional and agglutinative morphology, the lack of vowels in currents written texts, and the multiplicity of its forms; this induces a high level of lexical and syntactic ambiguity. It follows considerable difficulties for the automatic processing. The selection of a linguistic environment providing powerful tools and the ability to improve performance according to our needs has led us to use the platform language NooJ. We begin with a study followed by a large-coverage formalization of the Arabic lexicon. The built dictionary, baptised "El-DicAr" allows to link all the inflexional, morphological, syntactico-semantic information to the list of lemmas. Automatic inflexional and derivational routines applied to this list produce more than 3 million inflected forms. We propose a new finite state machine compiler that leads to an optimal storage through a combination of a sequential minimization algorithm and a dynamic compression routine for stored information. This dictionary acts as the linguistic engine for the automatic morpho-syntactic analyzer that we have developed. This analyzer includes a set of tools: a morphological analyzer that identifies the component morphemes of agglutinative forms using large coverage morphological grammars, a new algorithm for looking through finite-state transducers in order to deal with texts written in Arabic with regardless of their vocalisation statements, a corrector of the most frequent typographical errors, a named entities recognition tool based on a combination of the morphological analysis results and rules described into local grammar presented as Augmented Transition Networks ( ATNS), an automatic annotator and some tools for linguistic research and contextual exploration. In order to make our work available to the scientific community, we have developed an online concordance service “NooJ4Web: NooJ for the Web”. It provides instant results to different types of queries and displays statistical reports as well as the corresponding histograms. The listed services are offered in order to collect feedbacks and improve performance. This system is used to process Arabic, as well as French and English
Thollard, Franck. "Inférence grammaticale probabiliste pour l'apprentissage de la syntaxe en traitement de la langue naturelle." Saint-Etienne, 2000. http://www.theses.fr/2000STET4010.
Nguyen, Hong Quang. "Reconnaissance automatique de la parole continue : grand vocabulaire en vietnamien." Avignon, 2008. http://www.theses.fr/2008AVIG0155.
Development of the Vietnamese speech recognition has just started. This is due to the differences between Vietnamese language and Western languages, the speech recognition techniques broadly used for these languages (English, French for example. ) are not enough for developing directly a powerful Vietnamese speech recognition system. Taking into consideration the Vietnamese language characteristics in term of data (lexicon, language model) and model (tone model) representation should allow us to obtain promised results and better performances. The first difference is the semantic entities segmentation of the sentence. In Vietnamese, the word/concept consists of one or several syllables which are systematically separated by spaces (syllabic language). The segmentation of the sentence in words/concepts is an important stage for the isolating languages such as the Mandarin, the Cantonese, and the Thai but also for the Vietnamese. To improve the performance of automatic recognition system for Vietnamese, we built a polysyllabic word segmentation module for syllabic sentences. Two approaches were used: the first one uses a Vietnamese polysyllabic word dictionary whereas the second builds automatically this dictionary using the mutual information of the words as the grouping criterion, and a dynamic programming algorithm to simplify the treatments. The second difference is the crucial role of the tone in the Vietnamese language. The tone recognition is thus a fundamental aspect of the tonal language processing. In this thesis, we studied various methods to represent, in an optimal way, the fundamental frequency and the energy. We also were interested in finding a method to reduce the influence of the co-articulation phenomenon between tones. We furthermore used two approaches: an approach in frames by using hidden Markov models and a more general method based on the multi-level perceptrons. By integrating the processing of the linguistic (polysyllabic word lexicon) and acoustic (tone recognition) characteristics, the results were improved by practically 50 % (compared to the baseline system). These results prove that the addition of supplementary information, characteristics of Vietnamese language, improves considerably the performances of the speech recognition system
Hallé, Pierre. "Segmentation syllabique et reconnaissance des tons du chinois en parole continue." Paris 11, 1985. http://www.theses.fr/1985PA112204.
This is a preliminary study on phonetic decoding of a tone language: the Peking dialect. The choice of this language implies the use of strategies based on the syllable as the recognition unit. A new method of vowel nuclei detection is first proposed. This method combines the use of spectral variability and loudness. The second part is devoted to the recognition of tones, regardless of the segmentals on which they are carried. A normalization technique based on a general intonation curve estimate is first proposed. Then a probabilistic recognition technique is presented. We have called it "contrastive" because, in addition to using information on the intrinsic characteristics of tonal units, it also employs knowledge of the contrasts between these units
Yektaie, Mohammad-Hossein. "Réalisation d'un système adapté à la reconnaissance de caractères typographiques persans." La Rochelle, 1998. http://www.theses.fr/1998LAROS022.
Pham, Trong-Ton. "Modélisation et recherche de graphes visuels : une approche par modèles de langue pour la reconnaissance de scènes." Phd thesis, Université de Grenoble, 2010. http://tel.archives-ouvertes.fr/tel-00996067.
Mathey, Stéphanie. "Le rôle du voisinage orthographique lors de la reconnaissance visuelle des mots." Dijon, 1997. http://www.theses.fr/1997DIJOL004.
The aim of this research is to contribute to the understanding of lexical access mechanisms by manipulating orthographic neighborhood in lexical decision tasks. The results of experiments 1 and 2 show a facilitatory neighborhood frequency effect. This effect does not vary as a function of word size and may be due to a positive response bias induced by the task. The effect of orthographically similar words, neighbors by suppressing one letter, is examined in experiments 3 and 4. Identification performance for words with suppression neighbors and words with no neighbor are identical. We also replicate the facilitatory neighborhood frequency effect with traditional neighbors. In experiments 5, 6, and 7 we examine if a neighborhood relationship between the neighbors of a stimulus plays a role on the visual recognition of that stimulus. A facilitatory neighborhood distribution effect is observed, which is correctly predicted by the interactive-activation model. In addition, an inhibitory neighborhood frequency effect appears when the pseudoword context is difficult, which suggests a strategic bias. As a whole, the results are compatible with the interactive-activation model, but seem to vary according to the task difficulty. Explanations based on strategic effects are examined
Seng, Sopheap. "Vers une modélisation statistique multi-niveau du langage : application aux langues peu dotées." Phd thesis, Grenoble, 2010. http://www.theses.fr/2010GRENM010.
This PhD thesis focuses on the problems encountered when developing automatic speech recognition for under-resourced languages with a writing system without explicit separation between words. The specificity of the languages covered in our work requires automatic segmentation of text corpus into words in order to make the n-gram language modeling applicable. While the lack of text data has an impact on the performance of language model, the errors introduced by automatic segmentation can make these data even less usable. To deal with these problems, our research focuses primarily on language modeling, and in particular the choice of lexical and sub-lexical units, used by the recognition systems. We investigate the use of multiple units in speech recognition system. We validate these modeling approaches based on multiple units in recognition systems for a group of languages : Khmer, Vietnamese, Thai and Laotian
Seng, Sopheap. "Vers une modélisation statistique multi-niveau du langage : application aux langues peu dotées." Phd thesis, Université de Grenoble, 2010. http://tel.archives-ouvertes.fr/tel-00646236.
Segal, Natalia. "Analyse, représentation et modélisation de la prosodie pour la reconnaissance automatique de la parole." Paris 7, 2011. http://www.theses.fr/2011PA070041.
This thesis presents a new approach to automatic prosodic boundary and prosodic structure detection based on a theoretical hierarchical representation of prosodic organization of speech in French. We used a descriptive theory of the French prosodic System to create a rule based linguistic prosodic model suitable for the automatic treatment of spontaneous speech. This model allows finding automatically prosodic group boundaries and structuring them hierarchically. The prosodic structure of every phrase is thus represented in the form of a prosodic tree. This representation proved to be efficient for automatic processing of continuous speech in French. The resulting prosodic segmentation was compared to manual prosodic segmentation. Prosodic structure accuracy was also verified manually by an expert. We applied our model to different kinds of continuous spontaneous speech data with different phonemic and lexical segmentations: manual segmentation and different kinds of automatic segmentations. In particular, the application of our prosodic model to the output of a speech recognition System showed a satisfactory performance. There also bas been established a correlation between the level of the prosodic tree node and the boundary detection accuracy. Thus, it is possible to improve the precision of boundary detection by attributing a degree of confidence to the boundary according to its level in prosodic tree
Pagel, Vincent. "De l'utilisation d'informations acoustiques suprasegmentales en reconnaissance de la parole continue." Nancy 1, 1999. http://www.theses.fr/1999NAN10299.
Lembessi, Zacharoula-Pénélope. "Système de conjugaison, reconnaissance morphosyntaxique statistique lemmatisation automatique de la classe verbale du grec moderne standard." Université Marc Bloch (Strasbourg) (1971-2008), 2006. https://publication-theses.unistra.fr/public/theses_doctorat/2006/LEMBESSI_Penelope_2006.pdf.
In this dissertation we present the results as well as the methodology employed for the creation of an automatic (non-lexicon) morphosyntactic analysis and lemmatization system of the one-word Modern Greek Standard (MGS) verbal forms. . .
Liu, Li. "Modélisation pour la reconnaissance continue de la langue française parlée complétée à l'aide de méthodes avancées d'apprentissage automatique." Thesis, Université Grenoble Alpes (ComUE), 2018. http://www.theses.fr/2018GREAT057/document.
This PhD thesis deals with the automatic continuous Cued Speech (CS) recognition basedon the images of subjects without marking any artificial landmark. In order to realize thisobjective, we extract high level features of three information flows (lips, hand positions andshapes), and find an optimal approach to merging them for a robust CS recognition system.We first introduce a novel and powerful deep learning method based on the ConvolutionalNeural Networks (CNNs) for extracting the hand shape/lips features from raw images. Theadaptive background mixture models (ABMMs) are also applied to obtain the hand positionfeatures for the first time. Meanwhile, based on an advanced machine learning method Modi-fied Constrained Local Neural Fields (CLNF), we propose the Modified CLNF to extract theinner lips parameters (A and B ), as well as another method named adaptive ellipse model. Allthese methods make significant contributions to the feature extraction in CS. Then, due tothe asynchrony problem of three feature flows (i.e., lips, hand shape and hand position) in CS,the fusion of them is a challenging issue. In order to resolve it, we propose several approachesincluding feature-level and model-level fusion strategies combined with the context-dependentHMM. To achieve the CS recognition, we propose three tandem CNNs-HMM architectureswith different fusion types. All these architectures are evaluated on the corpus without anyartifice, and the CS recognition performance confirms the efficiency of our proposed methods.The result is comparable with the state of the art using the corpus with artifices. In parallel,we investigate a specific study about the temporal organization of hand movements in CS,especially about its temporal segmentation, and the evaluations confirm the superior perfor-mance of our methods. In summary, this PhD thesis applies the advanced machine learningmethods to computer vision, and the deep learning methodologies to CS recognition work,which make a significant step to the general automatic conversion problem of CS to sound.The future work will mainly focus on an end-to-end CNN-RNN system which incorporates alanguage model, and an attention mechanism for the multi-modal fusion
Echchourafi, Adnane. "Vers une reconnaissance des composés, pour une désambiguïsation automatique : composés à trois, quatre, cinq et six éléments." Université de Franche-Comté, 2006. http://www.theses.fr/2006BESA1010.
Nguyen, Viet Son. "Etudes de caractéristiques de la langue vietnamienne en vue de sa synthèse et de sa reconnaissance automatique : aspects statiques et dynamiques." Paris, Télécom ParisTech, 2009. http://www.theses.fr/2009ENST0056.
The work presented in this thesis aims to study profoundly the static and dynamic characteristics of the long vowels, short vowels, and the final stop consonants in the Vietnamese in calculating the rates of formant transitions. Two sets consonants vowels were analyzed: (C1) V1C2, and (C1)V1V2 in which C2 is one of the three final stop consonants / p, t, k / and V1 is one of the thirteen vowels (included nine long vowels and four short vowels). The objective of the analysis is to highlight the identification of specific characteristics which are then validated in the perception tests
Betari, Abdelkader. "Caractérisation des phonèmes de l'arabe standard en vue d'une reconnaissance automatique de la parole." Aix-Marseille 2, 1993. http://www.theses.fr/1993AIX22029.
Bendjillali, Mimoun. "La reconnaissance des langues et cultures d'origine : un facteur de réussite scolaire dans un contexte interculturel." Paris 8, 1996. http://www.theses.fr/1996PA081226.
Our ressearch deals with the education of children of algerian origin in the french elementary school. Involving many contradictions, school life the appears as a triple interaction between the teaching the native language and culture (ilcs), who imposes with his presence, his position and his segregated teaching, on the creation of intercultural approaches as a pedagogy
Jamborova-Lemay, Diana. "Analyse morphologique automatique du slovaque : étude approfondie du système linguistique slovaque et sa reconnaissance d'après la forme dans les textes scientifiques et techniques, application au machinisme agricole." Paris, INALCO, 2003. http://www.theses.fr/2003INAL0013.
Automatic morphological analysis of Slovak language is the first level of an automatical analyser for Slovak's scientifical and technical texts. Such a system could be used for different applications : automatic text indexation, automatic research of terminology or translation systems. A rule-based description of language's regularities as well as the use of all the formal level elements of words allow to reduce considerably the volume of dictionaries. Notably in case of inflectionally rich languages such as Slovak. The results obtained by our morphological analyser justify such an approach and confirm the high reliability of morphological analysis based on form-recognition for all lexical categories
Martin-Dastillung, Éveline. "Éléments pour un système de reconnaissance de contextes thématiques dans l'exploration d'un corpus textuel." Paris 4, 1992. http://www.theses.fr/1992PA040005.
This study is made in the field of computer assisted text analysis and exploration. Corpora data bases, (and here Frantext is no exception) generally go no further than accessing surface structure: the aim here is to outline a thematic lexicological knowledge-base allowing a partial exploration of deep structures by automatic recognition of conceptually associated words. The on-going data base described here will be regularly enriched with the lists obtained from analogy notes in dictionaries, from specific subject-oriented studies and more particularly from semi-automatized context selection made from constantly up-dated specific corpora. Sematic, stylistic and chronological criteria will be taken into account and incorporated into multiple choice menus which best match query and answer. The main obstacle, besides homography and polysemy, is identification technique. Bringing together pertinent correlates raises a number of methodological questions, especially on those relations that can go towards linking meanings. Coupled with a full-text exploration, the data base may be expected to produce collections of contexts (quotations, examples, attestations) centered on a given theme. A simulation made from the concept ennui ("boredom") gives an overview of the way this may be done
Riewpituk, Weera. "Reconnaissance de la parole par des mots isolés, appliquée à la langue thaie, prenant en compte les tons lexicaux." Paris 11, 1985. http://www.theses.fr/1985PA112102.
Within the context of this thesis, we have worked out a mono-speaker recognition system of isolated words in thai language, a language which is tonally monosyllabic, taking into respect the lexical tones. This system uses the global method and the dynamic programming technique to compare the input parameters with the parameters in the dictionnary. The first phase of the thesis was in working out a system of recognition of isolated words using the parameters of partial correlation (PARCOR). In the second phase, we used the fundamental frequency for words which are phonetically identical
Djoudi, Mahieddine. "Contribution à l'étude et à la reconnaissance automatique de la parole en Arabe standard." Nancy 1, 1991. http://docnum.univ-lorraine.fr/public/SCD_T_1991_0395_DJOUDI.pdf.
Gonzalez, Preciado Matilde. "Méthodes de vision par ordinateur pour la reconnaissance de gestes naturelles dans le contexte de lʼannotation en langue des signes." Phd thesis, Université Paul Sabatier - Toulouse III, 2012. http://tel.archives-ouvertes.fr/tel-00768440.
Boujelbane, Jarraya Rahma. "Traitements linguistiques pour la reconnaissance automatique de la parole appliquée à la langue arabe : de l'arabe standard vers l'arabe dialectal." Thesis, Aix-Marseille, 2015. http://www.theses.fr/2015AIXM4121.
The different dialects of the arabic language have a large phonological, morphological, lexical and syntactic variations when compared to the standard written arabic language called MSA (Modern Standard Arabic). Until recently, these dialects were presented only in their oral form and most of the existing resources for the Arabic language is limited to the Standard Arabic (MSA), leading to an abundance of tools for the automatic processing of this variety. Given the significant differences between the MSA and DA, the performance of these tools fall down when processing AD. This situation leads to a significant increase of the ambiguity in computational approaches of AD.This thesis is part of this framework by modeling the oral spoken in the Tunisian media. This data source contains a significant amount of Code Switching (CS) between the normative language MSA and the Dialect spoken in Tunisia (DT). The presence of the latter in a disorderly manner in the discourse poses a serious problem for NLP (Natural Language Processing) and makes this oral a less resourced language. However, the resources required to model this oral are almost nonexistent. Thus, the objective of this thesis is to fill this gap in order to build a language model dedicated to an automatic recognition system for the oral spoken in the Tunisian media. For this reason, we describe in this thesis a resource generation methodologyand we evaluate it relative to a language modeling task. The results obtained are encouraging
Brito, Marcilio de. "Réalisation d'un analyseur morpho-syntaxique pour la reconnaissance du syntagme nominal : utilisation des grammaires affixes." Lyon 1, 1991. http://www.theses.fr/1991LYO10153.
Lê, Viêt Bac. "Reconnaissance automatique de la parole pour des langues peu dotées." Université Joseph Fourier (Grenoble), 2006. http://www.theses.fr/2006GRE10061.
Nowadays, computers are heavily used to communicate via text and speech. Text processing tools, electronic dictionaries, and even more advanced systems like text-to-speech or dictation are readily available for several languages. There are however more than 6900 languages in the world and only a small number possess the resources required for implementation of Human Language Technologies (HLT). Thus, HLT are mostly concerned by languages for which large resources are available or which have suddenly become of interest because of the economic or political scene. On the contrary, languages from developing countries or minorities have been less worked on in the past years. One way of improving this "language divide" is do more research on portability of HLT for multilingual applications. Among HLT, we are particularly interested in Automatic Speech Recognition (ASR). Therefore, we are interested in new techniques and tools for rapid development of ASR systems for under-resourced languages or π-languages when only limited resources are available. These languages are typically spoken in developing countries, but can nevertheless have many speakers. In this work, we investigate Vietnamese and Khmer, which are respectively spoken by 67 million and 13 million people, but for which speech processing services do not exist at all. Firstly, given the statistical nature of the methods used in ASR, a large amount of resources (vocabularies, text corpora, transcribed speech corpora, phonetic dictionaries) is crucial for building an ASR system for a new language. Concerning text resources, a new methodology for fast text corpora acquisition for π-languages is proposed and applied to Vietnamese and Khmer. Some specific problems in text acquisition and text processing for π-languages such as text normalization, text segmentation, text filtering are resolved. For fast developing of text processing tools for a new π-language, an open source generic toolkit named CLIPS-Text-Tk was developed during this thesis. Secondly, for acoustic modeling, we address particularly the use of acoustic-phonetic unit similarities for multilingual acoustic models portability to new languages. Notably, an estimation method of the similarity between two phonemes is first proposed. Based on these phoneme similarities, some estimation methods for polyphone similarity and clustered polyphonic model similarity are investigated. For a new language, a source/target acoustic-phonetic unit mapping table can be constructed with these similarity measures. Then, clustered models in the target language are duplicated from the nearest clustered models in the source language and adapted with limited data to the target language. Results obtained for Vietnamese demonstrate the feasibility and efficiency of these methods. The proposal of grapheme-based acoustic modeling, which avoids building a pronunciation dictionary, is also investigated in our work. Finally, our whole methodology is applied to design a Khmer ASR system which leads to 70% word accuracy and which was developed in only five months
Derouault, Anne-Marie. "Modélisation d'une langue naturelle pour la désambiguation des chaînes phonétiques." Paris 7, 1985. http://www.theses.fr/1985PA077028.
Font, Noëlle. "Rôle de la langue dans l'accès au lexique chez les bilingues : influence de la proximité orthographique et sémantique interlangue sur la reconnaissance visuelle de mots." Montpellier 3, 2001. http://www.theses.fr/2001MON30015.
Lange, Marielle. "De l'orthographe à la prononciation: nature des processus de conversion graphème-phonème dans la reconnaissance des mots écrits." Doctoral thesis, Universite Libre de Bruxelles, 2000. http://hdl.handle.net/2013/ULB-DIPOT:oai:dipot.ulb.ac.be:2013/211741.
Nguyen, Viet Son. "Etude des caractéristiques de la langue vietnamienne en vue de sa synthèse et de sa reconnaissance automatique. Aspects statiques et dynamiques." Phd thesis, Telecom ParisTech, 2009. http://tel.archives-ouvertes.fr/tel-01064853.
Aparicio, Xavier. "Mécanismes cognitifs du changement de langue chez les multilingues : études comportementales et électrophysiologiques." Thesis, Montpellier 3, 2010. http://www.theses.fr/2010MON30078/document.
The present work focuses on language switching mechanisms during visual word recognition in French / English / Spanish trilinguals. First, we examined access to processing for each language, independently from the presentation of other languages. Then, our concern was to determine the different relations between languages and the underlying lexical organization inside the multilingual memory, as well as the activation of cognitive mechanisms allowing language switching. With this goal in mind, we performed three series of experiments to examine the influences between the languages, and the cognitive cost subsequent to a language switching during visual word recognition. We compare the processing of non-cognate words belonging to the three languages, manipulating the language switching situations and recording behavioral and electrophysiological data during lexical decision, semantic categorization and semantic translation priming. Results of these experiments highlight a general slowdown of processing consecutive to a language switch for all languages. Moreover, we observed a cognitive cost related to language switching for all the three languages, but bilaterally larger when it concerns the two non-dominant languages (L2 to L3 and L3 to L2). The recording data confirms the hypothesis of lexical representations integrated into a shared lexicon of multilingual memory. The results are interpreted in the light of the main models accounting for bilingual memory
Solbiac, Rodolphe. "Migration et multiculturalisme dans l'oeuvre de Neil Bissoondath : exploration d'une reconnaissance singulière." Antilles-Guyane, 2005. http://www.theses.fr/2005AGUY0128.
This dissertation explores Neil Bissoondath' s text and Canadian literary and social contexts of the 80's and early 90's for an understanding of his surprising acceptance among mainstrearn Canadian writers. At a time when it is difficult for visible minorities' writers to have their books published talked about by Canadian criticism, what in Bissoondath's work and in Canadian literary space results in his status as a mainstrearn Canadian writer. Our studies of space and characters search for the signs producing the effect of the work on readers and causing its reception. They reveal a vision of Caribbean and South America as hostile places, where multi-ethnic nations are doomed to failure, of Canada as a refuge where Anglo European humanistic values are threatened by immigrant confinement to community. Our unveiling of the articulation between the reception Bissoondath's work and the reform of multiculturalism reveals that his access to the mainstrearn results from two causes. First, Bissoondath benefits from the institutionalisation of multiculturalism in the field of literature that promotes works by visible minorities ' writers. Then, its reception derives from a convergence between the ideology developed in its text about multiculturalism, celebrating Anglo Canadian values, and the re-defining of Canadian citizenship, that give precedence to common Canadian values over what make Canadians different. This situation of Bisoondath's work, exceptional in the late 80's, becomes commonplace by the of the 90's as it's publishing coincided with the emergence of a new Canadian literary landscape in which distinctions between mainstrearn and minority writers are abolished
Doignon, Nadège. "Les unités de lecture dans la reconnaissance visuelle des mots : le paradigme des conjonctions illusoires." Dijon, 2003. http://www.theses.fr/2003DIJOL009.
The aim was to study the nature of sublexical units with the illusory conjunction paradigm. We focused on syllable units and orthographic redundancy. Our results showed an influence of the both types of information on reading unit perception. The syllable effect was attenuated when the orthographic properties were not congruent with the syllable segmentation. Other results were found : orthographic clues such as consonants give some indication to the constitution of units ; a syllable frequency effect suggesting a mental representation of syllable : reading units were perceived in the process of lexical access but also affected by lexical information ; results are similar for beginning and expert readers. To conclude, the experiments provided converging evidence that reading units are influenced by orthographic and phonological prelexical representations that correspond respectively to orthographic redundancy and syllable unit information, and lexical representations
Garçon, Laurie-Amandine. "Développement de langue électronique : étude de mélanges complexes et de bactéries." Thesis, Université Grenoble Alpes (ComUE), 2015. http://www.theses.fr/2015GREAI069/document.
The objective of this PhD thesis is to explore the potential applications of the electronic tongue, based on combinatorial cross-reactive receptors and surface plasmon resonance imaging, for analysis and discrimination of different complex mixtures and bacteria. In this work, various complex mixtures of different nature such as wine, beer, and milk (either animal-based or plant-based) are used. It has been demonstrated that the electronic tongue is capable of responding differently to theses samples with good selectivity. For each of them, it can generate characteristic continuous 2D profile and 3D image, based on which the differentiation and classification of the complex mixtures have been carried out. Furthermore, it has been illustrated that the electronic tongue is efficient for monitoring the deterioration of milk. In the second part of this thesis, the electronic tongue has been applied for detection and analysis of bacteria. At first, some fluidic parameters have been optimized due to the variable morphology of these complexes and large biological objects. Under optimized experimental conditions, the electronic tongue is effective for analysis of bacteria with the possibility for quantification. Thereafter, the electronic tongue has allowed for the discrimination of different bacteria according to their genus, species and strains based on continuous 2D profiles and 3D images
Grandjean, Anaël. "Reconnaissance de langage en temps réel sur automates cellulaires 2D." Thesis, Montpellier, 2016. http://www.theses.fr/2016MONTT331/document.
Cellular automata were introduced in the 50s by J. von Neumann and S. Ulamas an efficient way of modeling massively parallel computation. Many variations of the model can be considered such as varying the dimension of the computation space or the communication capabilities of the computing cells. In a cellular automaton each cell can communicate only with a finite number of other cells called its neighbors. My work focuses on the impact of the choice of the neighbors on the algorithmic properties of the model. My first goal was to generalize some classical properties of computation models to the widest possible class of neighborhoods, in particular I prove a linear speedup theorem for any two dimensional neighborhood. I then study the difference between the complexity classes defined by different neighborhoods, show the existence of neighborhoods defining incomparable classes, and some sets of neighborhoods defining identical classes. Finally, I also discuss the impact of the dimension of the automata on their computational power