Dissertations / Theses on the topic 'Système de question-réponse visuels'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 17 dissertations / theses for your research on the topic 'Système de question-réponse visuels.'
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.
Dancette, Corentin. "Shortcut Learning in Visual Question Answering." Electronic Thesis or Diss., Sorbonne université, 2023. http://www.theses.fr/2023SORUS073.
Full textThis thesis is focused on the task of VQA: it consists in answering textual questions about images. We investigate Shortcut Learning in this task: the literature reports the tendency of models to learn superficial correlations leading them to correct answers in most cases, but which can fail when encountering unusual input data. We first propose two methods to reduce shortcut learning on VQA. The first, which we call RUBi, consists of an additional loss to encourage the model to learn from the most difficult and less biased examples -- those which cannot be answered solely from the question. We then propose SCN, a model for the more specific task of visual counting, which incorporates architectural priors designed to make it more robust to distribution shifts. We then study the existence of multimodal shortcuts in the VQA dataset. We show that shortcuts are not only based on correlations between the question and the answer but can also involve image information. We design an evaluation benchmark to measure the robustness of models to multimodal shortcuts. We show that existing models are vulnerable to multimodal shortcut learning. The learning of those shortcuts is particularly harmful when models are evaluated in an out-of-distribution context. Therefore, it is important to evaluate the reliability of VQA models, i.e. We propose a method to improve their ability to abstain from answering when their confidence is too low. It consists of training an external ``selector'' model to predict the confidence of the VQA model. This selector is trained using a cross-validation-like scheme in order to avoid overfitting on the training set
Lerner, Paul. "Répondre aux questions visuelles à propos d'entités nommées." Electronic Thesis or Diss., université Paris-Saclay, 2023. http://www.theses.fr/2023UPASG074.
Full textThis thesis is positioned at the intersection of several research fields, Natural Language Processing, Information Retrieval (IR) and Computer Vision, which have unified around representation learning and pre-training methods. In this context, we have defined and studied a new multimodal task: Knowledge-based Visual Question Answering about Named Entities (KVQAE).In this context, we were particularly interested in cross-modal interactions and different ways of representing named entities. We also focused on data used to train and, more importantly, evaluate Question Answering systems through different metrics.More specifically, we proposed a dataset for this purpose, the first in KVQAE comprising various types of entities. We also defined an experimental framework for dealing with KVQAE in two stages through an unstructured knowledge base and identified IR as the main bottleneck of KVQAE, especially for questions about non-person entities. To improve the IR stage, we studied different multimodal fusion methods, which are pre-trained through an original task: the Multimodal Inverse Cloze Task. We found that these models leveraged a cross-modal interaction that we had not originally considered, and which may address the heterogeneity of visual representations of named entities. These results were strengthened by a study of the CLIP model, which allows this cross-modal interaction to be modeled directly. These experiments were carried out while staying aware of biases present in the dataset or evaluation metrics, especially of textual biases, which affect any multimodal task
Embarek, Mehdi. "Un système de question-réponse dans le domaine médical : le système Esculape." Phd thesis, Université Paris-Est, 2008. http://tel.archives-ouvertes.fr/tel-00432052.
Full textBenamara, Farah. "Webcoop : un système de question-réponse coopératif sur le web." Toulouse 3, 2004. http://www.theses.fr/2004TOU30169.
Full textThis thesis describes the WEBCOOP system that aims at providing cooperative responses in French to natural language queries on the web. The main objectives of the system are : -the integration of reasoning procedures with a variety of knowledge bases as well as real life data extracted from web pages in order to produce web style natural language responses. -major and new feature: the integration of a cooperative know-how component that goes beyond the mere recognition of a user misconception
Moriceau, Véronique. "Intégration de données dans un système question-réponse sur le Web." Toulouse 3, 2007. http://www.theses.fr/2007TOU30019.
Full textIn the framework of question-answering systems on the Web, our main goals are to model, develop and evaluate a system which can, from a question in natural language, search for relevant answers on the Web and generate a synthetic answer, even if the search engine selected several candidate answers. We focused on temporal and numerical questions. Our system deals with : - the integration of data from candidate answers by using a knowledge base and knowledge extracted from the Web. This component allows the detection of data inconsistencies and deals with user expectations in order to produce a relevant answer, - the generation of synthetic answers in natural language which are relevant w. R. T users. Indeed, generated answers have to be short, understandable and have to express the cooperative know-how which has been used to solve data inconsistencies. We also propose evaluation methods to evaluate our system from a technical and cognitive point of view
Sedogbo, Célestin. "De la grammaire en chaîne du français à un système question-réponse." Aix-Marseille 2, 1987. http://www.theses.fr/1987AIX22092.
Full textBernard, Guillaume. "Réordonnancement de candidats reponses pour un système de questions-réponses." Phd thesis, Université Paris Sud - Paris XI, 2011. http://tel.archives-ouvertes.fr/tel-00606025.
Full textNicaud, Lydia. "Le raisonnement caricatural : un guide pour le raisonnement dans un système question-réponse en langage naturel." Paris 11, 1986. http://www.theses.fr/1986PA112075.
Full textQuestion answering systems in natural language, using a lot of production rules, have some difficulties to choice the relevant rules, to determine the goal to reach, without inducing combinatory explosion. This thesis proposes a reasoning strategy which works upon rude knowledge, in order to blaze trails and to provide a guiding for a natural language reasoned. This strategy is also able to pick up a few number of relevant rules, and to select relevant goals
Monceaux, Laura. "Adaptation du niveau d'analyse des interventions dans un dialogue : application à un système de question-réponse." Paris 11, 2002. http://www.theses.fr/2002PA112291.
Full textDue to the variety of dialogue types, we studied how to adapt, in a generic way, the level of user interventions analysis. An analysis by keywords, recommended by systems of specific dialogues (relative to a particular task) is inadequate to handle such interventions, because therefore, it is impossible to represent all the world knowledge and solve the conflicts arising from this variety. We developed an analysis of the interventions, independent from their domains and based on the syntax interventions. In so doing, we were confronted with the choice of syntactic analyzer. To solve it, we studied the various existing syntactic parsers by constructing a classification according to their capacities, followed by the development of an evaluation protocol of these analyzers for French. Further to this evaluation, it appeared interesting to develop an algorithm of compromise between several analyses to return the most plausible analysis. This will allow us not only to use the capacities of every analyzes but also to quantify every information returned by a confidence rate. From the intervention's syntax and the semantic knowledge provided by the lexical base WordNet (synonym, hyperonym), we developed a system to extract the intervention's intention and its propositional contents. Particularly, we were interested in the question interventions : the propositional contents rely upon the extraction of the answer type and of the object of the question. To estimate the efficiency of these criteria, this analysis was integrated into the question-answering system developed in the LIR group
Saneifar, Hassan. "Locating Information in Heterogeneous log files." Thesis, Montpellier 2, 2011. http://www.theses.fr/2011MON20092/document.
Full textIn this thesis, we present contributions to the challenging issues which are encounteredin question answering and locating information in complex textual data, like log files. Question answering systems (QAS) aim to find a relevant fragment of a document which could be regarded as the best possible concise answer for a question given by a user. In this work, we are looking to propose a complete solution to locate information in a special kind of textual data, i.e., log files generated by EDA design tools.Nowadays, in many application areas, modern computing systems are instrumented to generate huge reports about occurring events in the format of log files. Log files are generated in every computing field to report the status of systems, products, or even causes of problems that can occur. Log files may also include data about critical parameters, sensor outputs, or a combination of those. Analyzing log files, as an attractive approach for automatic system management and monitoring, has been enjoying a growing amount of attention [Li et al., 2005]. Although the process of generating log files is quite simple and straightforward, log file analysis could be a tremendous task that requires enormous computational resources, long time and sophisticated procedures [Valdman, 2004]. Indeed, there are many kinds of log files generated in some application domains which are not systematically exploited in an efficient way because of their special characteristics. In this thesis, we are mainly interested in log files generated by Electronic Design Automation (EDA) systems. Electronic design automation is a category of software tools for designing electronic systems such as printed circuit boards and Integrated Circuits (IC). In this domain, to ensure the design quality, there are some quality check rules which should be verified. Verification of these rules is principally performed by analyzing the generated log files. In the case of large designs that the design tools may generate megabytes or gigabytes of log files each day, the problem is to wade through all of this data to locate the critical information we need to verify the quality check rules. These log files typically include a substantial amount of data. Accordingly, manually locating information is a tedious and cumbersome process. Furthermore, the particular characteristics of log files, specially those generated by EDA design tools, rise significant challenges in retrieval of information from the log files. The specific features of log files limit the usefulness of manual analysis techniques and static methods. Automated analysis of such logs is complex due to their heterogeneous and evolving structures and the large non-fixed vocabulary.In this thesis, by each contribution, we answer to questions raised in this work due to the data specificities or domain requirements. We investigate throughout this work the main concern "how the specificities of log files can influence the information extraction and natural language processing methods?". In this context, a key challenge is to provide approaches that take the log file specificities into account while considering the issues which are specific to QA in restricted domains. We present different contributions as below:> Proposing a novel method to recognize and identify the logical units in the log files to perform a segmentation according to their structure. We thus propose a method to characterize complex logicalunits found in log files according to their syntactic characteristics. Within this approach, we propose an original type of descriptor to model the textual structure and layout of text documents.> Proposing an approach to locate the requested information in the log files based on passage retrieval. To improve the performance of passage retrieval, we propose a novel query expansion approach to adapt an initial query to all types of corresponding log files and overcome the difficulties like mismatch vocabularies. Our query expansion approach relies on two relevance feedback steps. In the first one, we determine the explicit relevance feedback by identifying the context of questions. The second phase consists of a novel type of pseudo relevance feedback. Our method is based on a new term weighting function, called TRQ (Term Relatedness to Query), introduced in this work, which gives a score to terms of corpus according to their relatedness to the query. We also investigate how to apply our query expansion approach to documents from general domains.> Studying the use of morpho-syntactic knowledge in our approaches. For this purpose, we are interested in the extraction of terminology in the log files. Thus, we here introduce our approach, named Exterlog (EXtraction of TERminology from LOGs), to extract the terminology of log files. To evaluate the extracted terms and choose the most relevant ones, we propose a candidate term evaluation method using a measure, based on the Web and combined with statistical measures, taking into account the context of log files
Soumana, Ibrahim. "Interrogation des sources de données hétérogènes : une approche pour l'analyse des requêtes." Thesis, Besançon, 2014. http://www.theses.fr/2014BESA1015/document.
Full textNo english summary available
Elbaz, Ilan. "Un système de question-réponse simple appliqué à SQuAD." Thesis, 2020. http://hdl.handle.net/1866/24313.
Full textThe Question-Answering task (QA) is a well established Natural Language Processing (NLP) task. Generally speaking, it consists in answering questions using documents (textual or otherwise) or conversations, making use of knowledge if necessary and implementing inference mechanisms. Thus, depending on the data set and the task associated with it, the system must be able to detect and understand the useful elements to correctly answer each of the questions asked. A lot of progress has been made in recent years with increasingly complex neural models. They are however expensive in production, and relatively opaque. Due to this opacity, it is diÿcult to accurately predict the behavior of some models and thus, to predict when these systems will return wrong answers. Unlike the vast majority of systems currently proposed, in this thesis we will try to solve this task with models with controllable size. We will focus mainly on feature-based approaches. The goal in restricting the size of the models is that they generalize better. So we will measure what these models capture in order to assess the granularity of their "understanding" of the language. Also, by analyzing the gaps of controllable size models, we will be able to highlight what more complex models have captured. To carry out our study, we evaluate ourselves here on SQuAD: a popular data set o˙ered by Standford University.
Merdaoui, Badis. "QUERI : un système de question-réponse collaboratif et interactif." Thèse, 2005. http://hdl.handle.net/1866/16700.
Full textBélanger, Luc. "Architecture question-réponse pour l'automatisation des services d'information." Thèse, 2006. http://hdl.handle.net/1866/16724.
Full textLazzouni, Latifa L. "Réponse auditive oscillatoire chez le non-voyant : investigation par magnétoencéphalographie." Thèse, 2012. http://hdl.handle.net/1866/8717.
Full textBlind persons show in their everyday life that they can efficiently adapt to visual deprivation by relying on their spared senses like touch or the sense of hearing. They also show they can challenge their environment without vision and sometimes even demonstrate superior abilities compared to sighted counterparts. In the last decades, research got more interested in adaptive capabilities of the blinds especially with the advent of new imaging techniques which made it possible to make giant steps investigating new avenues in the field of brain plasticity after sensory loss. The superior abilities of blind individuals take the form of a more efficient use of auditory and tactile information and find their neuronal correlates in the deafferented visual cortex. The visual cortex of the blind is still highly functional after visual deprivation and is recruited for the processing of cross modal auditory and tactile stimulations. It can even show implication in higher level memory or language processes. This functional involvement results from the plasticity of the visual cortex which is its ability to change its structure, its function and to adapt its interactions with the other systems in the absence of vision. Cortical plasticity is not exclusive to the visual cortex of the blind but is a permanent state of the brain. To appreciate cortical activity in the visual cortex of blind individuals, a measure of excitability of its neurons is used. This measure is represented by the recovery of the N1 component in ERPs to target detection, which is shorter in the auditory modality for the blind. Evoked potentials and evoked fields components in EEG and MEG have been shown to be reorganized in favour of the visual cortex of blind individuals compared to sighted ones for the auditory and tactile modalities. Posterior location for such components was found in the blind. The auditory steady-state response is another brain response that received less interest in the study of cortical reorganization after sensory loss. The ASSR has the advantage of oscillating at the stimulation rhythm and is characterized by a response in the auditory cortices tagged to the stimulation frequencies. The tag takes the form of an important spectral energy peak at the frequencies of stimulation in auditory areas. The ASSR is localized in left and right primary auditory areas, with this regard any posterior shift in the location of source activity in blind individuals also tagged to stimulation frequencies would be considered as an evidence of functional reorganization following sensory deprivation. The objectives of this work are to make use of the characteristics of the ASSR to amplitude modulated tones (AM) to investigate neural correlates of cross modal functional reorganization in the visual cortex of the blind for the processing of AM tones. The first study is a validation of the frequency tagging paradigm. A change detection auditory task can modulate the envelope amplitude of the ASSR response. The same paradigm is used to investigate cross modal reorganisation after long and short term visual deprivation. In this first study a group of healthy sighted individuals detected a change in the carrier frequency of AM tones, with eyes opened during monaural and dichotic listening. Two conditions were tested an active condition where they had to press a button each time they hear the change and a passive condition. Results show a significant increase in the envelope amplitude of the ASSR to the onset of the carrier frequency change, only for dichotic presentation. Patterns of activations of the ASSR were maintained, with larger responses in the hemisphere contralateral to the stimulated ear and binaural suppression for the ipsilateral inputs for the dichotic presentation. The second study was aimed to show that rapid changes in the ASSR to amplitude modulated tones (MA) are possible after short term sensory deprivation, by blindfolding sighted individuals for six hours. The same detection task was used but not the passive condition. Results show a modulation of the dichotic response in visual areas. The occipital source activity found, showed an auditory property as a binaural beat, which means an oscillating ASSR at a frequency equal to the difference of the frequencies presented to each ear. This effect was present in half of the participants and took place at the end of the blindfolding time. Cortical representation of the occipital sources showed a displacement of source activities in the antero-posterior direction at the end of transitory deprivation period. In the third study we compared the ASSR processing between early blind individuals (congenitally blind) group and healthy sighted controls group, to investigate the neural correlates of functional reorganization of this response after long term visual deprivation. Results show significant differences in the spectral representation of the response between the two groups. Important auditory temporal activations were found in the two groups. Distributed sources were localized in primary and secondary auditory areas for the two groups. A difference was found in blind individuals who showed additional activations of inferior temporal areas, known to be activated by objects vision in sighted individuals and being part of the what visual pathway. The results presented here are in line with a rapid reorganization of the ASSR after short term visual deprivation, and the implication of visual areas in the processing of AM tones for long term sensory deprivation in the congenitally blind. This was made possible by the unmasking of existing connections between auditory and visual cortices. Long term deprivation leads to plastic changes, in the auditory modality as a first step by the extension of activity to superior and middle temporal areas, then to cross modal changes with the functional involvement of inferior temporal areas in the processing of AM tones, considered as visual objects. This reorganization is likely to be mediated through lateral cortico-cortical connections.
Bouneffouf, Djallel. "Rôle de l'inférence temporelle dans la reconnaissance de l'inférence textuelle." Phd thesis, 2008. http://tel.archives-ouvertes.fr/tel-00786827.
Full textAdam, Pierre. "Améliorations d'artefacts sur panneaux LCD." Phd thesis, 2008. http://tel.archives-ouvertes.fr/tel-00396368.
Full text