Das GRADE-System: ein internationaler Ansatz zur Vereinheitlichung der Graduierung von Evidenz und Empfehlungen in Leitlinien. [The GRADE System: an international approach to standardize the graduation of evidence and recommendations in guidelines]

Clinical practice guidelines have become an important source of information to support clinicians in the management of individual patients. However, current guideline methods have limitations that include the lack of separating the quality of evidence from the strength of recommendations. The Grading of Recommendations, Assessment, Development and Evaluation (GRADE) working group, an international collaboration of guideline developers, methodologists, and clinicians have developed a system that addresses these shortcomings. Core elements include transparent methodology for grading the quality of evidence, the distinction between quality of the evidence and strength of a recommendation, an explicit balancing of benefits and harms of health care interventions, an explicit recognition of the values and preferences that underlie recommendations. The GRADE system has been piloted in various practice settings to ensure that it captures the complexity involved in evidence assessment and grading recommendations while maintaining simplicity and practicality. Many guideline organizations and medical societies have endorsed the system and adopted it for their guideline processes.

The use of positional scanning synthetic combinatorial libraries (PS-SCL) to study T Lympohcyte specificity and degeneracy

Les cellules CD8? T cytolytiques (CTL) sont les principaux effecteurs du système immunitaire adaptatif contre les infections et les tumeurs. La récente identification d?antigènes tumoraux humains reconnus par des cellules T cytolytiques est la base pour le, développement des vaccins antigène spécifiques contre le cancer. Le nombre d?antigènes tumoraux reconnus par des CTL que puisse être utilisé comme cible pour la vaccination des patients atteints du cancer est encore limité. Une nouvelle technique, simple et rapide, vient d?être proposée pour l?identification d?antigènes reconnus par des CTL. Elle se base sur l?utilisation de librairies combinatoriales de peptides arrangées en un format de "scanning" ou balayage par position (PS-SCL). La première partie de cette étude a consisté à valider cette nouvelle technique par une analyse détaillée de la reconnaissance des PS-SCL par différents clones de CTL spécifiques pour des antigènes associés à la tumeur (TAA) connus ainsi que par des clones de spécificité inconnue. Les résultats de ces analyses révèlent que pour tous les clones, la plupart des acides aminés qui composent la séquence du peptide antigénique naturel ont été identifiés par l?utilisation des PS-SCL. Les résultats obtenus ont permis d?identifier des peptides analogues ayant une antigènicité augmentée par rapport au peptide naturel, ainsi que des peptides comportant de multiples modifications de séquence, mais présentant la même réactivité que le peptide naturel. La deuxième partie de cette étude a consisté à effectuer des analyses biométriques des résultats complexes générés par la PS-SCL. Cette approche a permis l?identification des séquences correspondant aux épitopes naturels à partir de bases de données de peptides publiques. Parmi des milliers de peptides, les séquences naturelles se trouvent comprises dans les 30 séquences ayant les scores potentiels de stimulation les plus élevés pour chaque TAA étudié. Mais plus important encore, l?utilisation des PS-SCL avec un clone réactif contre des cellules tumorales mais de spécificité inconnue nous a permis d?identifier I?epitope reconnu par ce clone. Les données présentées ici encouragent l?utilisation des PS-SCL pour l?identification et l?optimisation d?épitopes pour des CTL réactifs anti-tumoraux, ainsi que pour l?étude de la reconnaissance dégénérée d?antigènes par les CTL.<br/><br/>CD8+ cytolytic T lymphocytes (CTL) are the main effector cells of the adaptive immune system against infection and tumors. The recent identification of moleculariy defined human tumor Ags recognized by autologous CTL has opened new opportunities for the development of Ag-specific cancer vaccines. Despite extensive work, however, the number of CTL-defined tumor Ags that are suitable targets for the vaccination of cancer patients is still limited, especially because of the laborious and time consuming nature of the procedures currentiy used for their identification. The use of combinatorial peptide libraries in positionai scanning format (Positional Scanning Synthetic Combinatorial Libraries, PS-SCL)' has recently been proposed as an alternative approach for the identification of these epitopes. To validate this approach, we analyzed in detail the recognition of PS-SCL by tumor-reactive CTL clones specific for multiple well-defined tumor-associated Ags (TAA) as well as by tumor-reactive CTL clones of unknown specificity. The results of these analyses revealed that for all the TAA-specific clones studied most of the amino acids composing the native antigenic peptide sequences could be identified through the use of PS-SCL. Based on the data obtained from the screening of PS-SCL, we could design peptide analogs of increased antigenicity as well as cross-reactive analog peptides containing multiple amino acid substitutions. In addition, the resuits of PS-SCL-screening combined with a recently developed biometric data analysis (PS-SCL-based biometric database analysis) allowed the identification of the native peptides in public protein databases among the 30 most active sequences, and this was the case for all the TAA studied. More importantiy, the screening of PS- SCL with a tumor-reactive CTL clone of unknown specificity resulted in the identification of the actual epitope. Overall, these data encourage the use of PS-SCL not oniy for the identification and optimization of tumor-associated CTL epitopes, but also for the analysis of degeneracy in T lymphocyte receptor (TCR) recognition of tumor Ags.<br/><br/>Les cellules T CD8? cytolytiques font partie des globules blancs du sang et sont les principales responsables de la lutte contre les infections et les tumeurs. Les immunologistes cherchent depuis des années à identifier des molécules exprimées et présentées à la surface des tumeurs qui puissent être reconnues par des cellules T CD8? cytolytiques capables ensuite de tuer ces tumeurs de façon spécifique. Ce type de molécules représente la base pour le développement de vaccins contre le cancer puisqu?elles pourraient être injectées aux patients afin d?induire une réponse anti- tumorale. A présent, il y a très peu de molécules capables de stimuler le système immunitaire contre les tumeurs qui sont connues parce que les techniques développées à ce jour pour leur identification sont complexes et longues. Une nouvelle technique vient d?être proposée pour l?identification de ce type de molécules qui se base sur l?utilisation de librairies de peptides. Ces librairies représentent toutes les combinaisons possibles des composants de base des molécules recherchées. La première partie de cette étude a consisté à valider cette nouvelle technique en utilisant des cellules T CD8? cytolytiques capables de tuer des cellules tumorales en reconnaissant une molécule connue présente à leur surface. On a démontré que l?utilisation des librairies permet d?identifier la plupart des composants de base de la molécule reconnue par les cellules T CD8? cytolytiques utilisées. La deuxième partie de cette étude a consisté à effectuer une recherche des molécules potentiellement actives dans des protéines présentes dans des bases des données en utilisant un programme informatique qui permet de classer les molécules sur la base de leur activité biologique. Parmi des milliers de molécules de la base de données, celles reconnues par nos cellules T CD8? cytolytiques ont été trouvées parmi les plus actives. Plus intéressant encore, la combinaison de ces deux techniques nous a permis d?identifier la molécule reconnue par une population de cellules T CD8? cytolytiques ayant une activité anti-tumorale, mais pour laquelle on ne connaissait pas la spécificité. Nos résultats encouragent l?utilisation des librairies pour trouver et optimiser des molécules reconnues spécifiquement par des cellules T CD8? cytolytiques capables de tuer des tumeurs.

Speaker recognition improvement using blind inversion of distortions

In this paper we propose the inversion of nonlinear distortions in order to improve the recognition rates of a speaker recognizer system. We study the effect of saturations on the test signals, trying to take into account real situations where the training material has been recorded in a controlled situation but the testing signals present some mismatch with the input signal level (saturations). The experimental results shows that a combination of several strategies can improve the recognition rates with saturated test sentences from 80% to 89.39%, while the results with clean speech (without saturation) is 87.76% for one microphone.

NBAS mutations cause a multisystem disorder involving bone, connective tissue, liver, immune system, and retina.

We report two unrelated patients with a multisystem disease involving liver, eye, immune system, connective tissue, and bone, caused by biallelic mutations in the neuroblastoma amplified sequence (NBAS) gene. Both presented as infants with recurrent episodes triggered by fever with vomiting, dehydration, and elevated transaminases. They had frequent infections, hypogammaglobulinemia, reduced natural killer cells, and the Pelger-Huët anomaly of their granulocytes. Their facial features were similar with a pointed chin and proptosis; loose skin and reduced subcutaneous fat gave them a progeroid appearance. Skeletal features included short stature, slender bones, epiphyseal dysplasia with multiple phalangeal pseudo-epiphyses, and small C1-C2 vertebrae causing cervical instability and myelopathy. Retinal dystrophy and optic atrophy were present in one patient. NBAS is a component of the synthaxin-18 complex and is involved in nonsense-mediated mRNA decay control. Putative loss-of-function mutations in NBAS are already known to cause disease in humans. A specific founder mutation has been associated with short stature, optic nerve atrophy and Pelger-Huët anomaly of granulocytes (SOPH) in the Siberian Yakut population. A more recent report associates NBAS mutations with recurrent acute liver failure in infancy in a group of patients of European descent. Our observations indicate that the phenotypic spectrum of NBAS deficiency is wider than previously known and includes skeletal, hepatic, metabolic, and immunologic aspects. Early recognition of the skeletal phenotype is important for preventive management of cervical instability. © 2015 Wiley Periodicals, Inc.

Time to identify cardiac arrest and provide dispatch-assisted cardio-pulmonary resuscitation in a criteria-based dispatch system.

INTRODUCTION: Dispatch-assisted cardiopulmonary resuscitation (DA-CPR) plays a key role in out-of-hospital cardiac arrests. We sought to measure dispatchers' performances in a criteria-based system in recognizing cardiac arrest and delivering DA-CPR. Our secondary purpose was to identify the factors that hampered dispatchers' identification of cardiac arrests, the factors that prevented them from proposing DA-CPR, and the factors that prevented bystanders from performing CPR. METHODS AND RESULTS: We reviewed dispatch recordings for 1254 out-of-hospital cardiac arrests occurring between January 1, 2011 and December 31, 2013. Dispatchers correctly identified cardiac arrests in 71% of the reviewed cases and 84% of the cases in which they were able to assess for patient consciousness and breathing. The median time to recognition of the arrest was 60s. The median time to start chest compression was 220s. CONCLUSIONS: This study demonstrates that performances from a criteria-based dispatch system can be similar to those from a medical-priority dispatch system regarding out-of-hospital cardiac arrest (OHCA) time recognition and DA-CPR delivery. Agonal breathing recognition remains the weakest link in this sensitive task in both systems. It is of prime importance that all dispatch centers tend not only to implement DA-CPR but also to have tools to help them reach this objective, as today it should be mandatory to offer this service to the community. In order to improve benchmarking opportunities, we completed previously proposed performance standards as propositions.

Optical Character Recognition for Symbian OS

This Master's thesis addresses the design and implementation of the optical character recognition (OCR) system for a mobile device working on the Symbian operating system. The developed OCR system, named OCRCapriccio, emphasizes the modularity, effective extensibility and reuse. The system consists of two parts which are the graphical user interface and the OCR engine that was implemented as a plug-in. In fact, the plug-in includes two implementations of the OCR engine for enabling two types of recognition: the bitmap comparison based recognition and statistical recognition. The implementation results have shown that the approach based on bitmap comparison is more suitable for the Symbian environment because of its nature. Although the current implementation of bitmap comparison is lacking in accuracy, further development should be done in its direction. The biggest challenges of this work were related to developing an OCR scheme that would be suitable for Symbian OS Smartphones that have limited computational power and restricted resources.

Creating dynamic continuous improvement system for Larox Corporation

Today´s organizations must have the ability to react to rapid changes in the market. These rapid changes cause pressure to continuously find new efficient ways to organize work practices. Increased competition requires businesses to become more effective and to pay attention to quality of management and to make people to understand their work's impact on the final result. The fundamentals in continmuois improvement are systematic and agile tackling of indentified individual process constraints and the fact tha nothin finally improves without changes. Successful continuous improvement requires management commitment, education, implementation, measurement, recognition and regeneration. These ingredients form the foundation, both for breakthrough projects and small step ongoing improvement activities. One part of the organization's management system are the quality tools, which provide systematic methodologies for identifying problems, defining their root causes, finding solutions, gathering and sorting of data, supporting decision making and implementing the changes, and many other management tasks. Organizational change management includes processes and tools for managing the people in an organizational level change. These tools include a structured approach, which can be used for effective transition of organizations through change. When combined with the understanding of change management of individuals, these tools provide a framework for managing people in change,

System-level characterization of Th2 cell development and immune cell responses to ZnO and TiO2 nanoparticles

Asthma and allergy are common diseases and their prevalence is increasing. One of the hypotheses that explains this trend is exposure to inhalable chemicals such as traffi c-related air pollution. Epidemiological research supports this theory, as a correlation between environmental chemicals and allergic respiratory diseases has been found. In addition to ambient airborne particles, one may be exposed to engineered nanosized materials that are actively produced due to their favorable physico-chemical properties compared to their bulk size counterparts. On the cellular level, improper activity of T helper (Th) cells has been connected to allergic reactions. Th cells can differentiate into functionally different effector subsets, which are identifi ed according to their characteristic cytokine profi les resulting in specifi c ability to communicate with other cells. Th2 cells activate humoral immunity and stimulate eradication of extracellular pathogens. However, persistent predominance of Th2 cells is involved in a development of number of allergic diseases. The cytokine environment at the time of antigen recognition is the major factor determining the polarization of a naïve Th cell. Th2 cell differentiation is initiated by IL4, which signals via transcription factor STAT6. Although the importance of this pathway has been evaluated in the mouse studies, the signaling components involved have been largely unknown. The aim of this thesis was to identify molecules, which are under the control of IL4 and STAT6 in Th cells. This was done by using system-level analysis of STAT6 target genes at genome, mRNA and protein level resulting in identifi cation of various genes previously not connected to Th2 cell phenotype acquisition. In the study, STAT6-mediated primary and secondary target genes were dissection from each other and a detailed transcriptional kinetics of Th2 cell polarization of naïve human CD4+ T cells was collected. Integration of these data revealed the hierarchy of molecular events that mediates the differentiation towards Th2 cell phenotype. In addition, the results highlighted the importance of exploiting proteomics tools to complement the studies on STAT6 target genes identifi ed through transcriptional profi ling. In the last subproject, the effects of the exposure with ZnO and TiO2 nanoparticles was analyzed in Jurkat T cell line and in primary human monocyte-derived macrophages and dendritic cells to evaluate their toxicity and potential to cause infl ammation. Identifi cation of ZnO-derived gene expression showed that the same nanoparticles may elicit markedly distinctive responses in different cell types, thus underscoring the need for unbiased profi ling of target genes and pathways affected. The results gave additional proof that the cellular response to nanosized ZnO is due to leached Zn2+ ions. The approach used in ZnO and TiO2 nanoparticle study demonstrated the value of assessing nanoparticle responses through a toxicogenomics approach. The increased knowledge of Th2 cell signaling will hopefully reveal new therapeutic nodes and eventually improve our possibilities to prevent and tackle allergic infl ammatory diseases.

PDM system functions and utilizations analysis to improve efficiency of sheet metal product design and manufacturing

This study will concentrate on Product Data Management (PDM) systems, and sheet metal design features and classification. In this thesis, PDM is seen as an individual system which handles all product-related data and information. The meaning of relevant data is to take the manufacturing process further with fewer errors. The features of sheet metals are giving more information and value to the designed models. The possibility of implementing PDM and sheet metal features recognition are the core of this study. Their integration should make the design process faster and manufacturing-friendly products easier to design. The triangulation method is the basis for this research. The sections of this triangle are: scientific literature review, interview using the Delphi method and the author’s experience and observations. The main key findings of this study are: (1) the area of focus in triangle (the triangle of three different point of views: business, information exchange and technical) depends on the person’s background and their role in the company, (2) the classification in the PDM system (and also in the CAD system) should be done using the materials, tools and machines that are in use in the company and (3) the design process has to be more effective because of the increase of industrial production, sheet metal blank production and the designer’s time spent on actual design and (4) because Design For Manufacture (DFM) integration can be done with CAD-programs, DFM integration with the PDM system should also be possible.

MHC-restricted antigen presentation and recognition: constraints on gene, recombinant and peptide vaccines in humans

The target of any immunization is to activate and expand lymphocyte clones with the desired recognition specificity and the necessary effector functions. In gene, recombinant and peptide vaccines, the immunogen is a single protein or a small assembly of epitopes from antigenic proteins. Since most immune responses against protein and peptide antigens are T-cell dependent, the molecular target of such vaccines is to generate at least 50-100 complexes between MHC molecule and the antigenic peptide per antigen-presenting cell, sensitizing a T cell population of appropriate clonal size and effector characteristics. Thus, the immunobiology of antigen recognition by T cells must be taken into account when designing new generation peptide- or gene-based vaccines. Since T cell recognition is MHC-restricted, and given the wide polymorphism of the different MHC molecules, distinct epitopes may be recognized by different individuals in the population. Therefore, the issue of whether immunization will be effective in inducing a protective immune response, covering the entire target population, becomes an important question. Many pathogens have evolved molecular mechanisms to escape recognition by the immune system by variation of antigenic protein sequences. In this short review, we will discuss the several concepts related to selection of amino acid sequences to be included in DNA and peptide vaccines.

The conservative physiology of the immune system

Current immunological opinion disdains the necessity to define global interconnections between lymphocytes and regards natural autoantibodies and autoreactive T cells as intrinsically pathogenic. Immunological theories address the recognition of foreignness by independent clones of lymphocytes, not the relations among lymphocytes or between lymphocytes and the organism. However, although extremely variable in cellular/molecular composition, the immune system preserves as invariant a set of essential relations among its components and constantly enacts contacts with the organism of which it is a component. These invariant relations are reflected, for example, in the life-long stability of profiles of reactivity of immunoglobulins formed by normal organisms (natural antibodies). Oral contacts with dietary proteins and the intestinal microbiota also result in steady states that lack the progressive quality of secondary-type reactivity. Autoreactivity (natural autoantibody and autoreactive T cell formation) is also stable and lacks the progressive quality of clonal expansion. Specific immune responses, currently regarded as the fundament of the operation of the immune system, may actually result from transient interruptions in this stable connectivity among lymphocytes. More permanent deficits in interconnectivity result in oligoclonal expansions of T lymphocytes, as seen in Omenn's syndrome and in the experimental transplantation of a suboptimal diversity of syngeneic T cells to immunodeficient hosts, which also have pathogenic consequences. Contrary to theories that forbid autoreactivity as potentially pathogenic, the physiology of the immune system is conservative and autoreactive. Pathology derives from failures of these conservative mechanisms.

Semantic and fuzzy modelling for human behaviour recognition in smart spaces : a case study on ambient assisted living.

Human activity recognition in everyday environments is a critical, but challenging task in Ambient Intelligence applications to achieve proper Ambient Assisted Living, and key challenges still remain to be dealt with to realize robust methods. One of the major limitations of the Ambient Intelligence systems today is the lack of semantic models of those activities on the environment, so that the system can recognize the speci c activity being performed by the user(s) and act accordingly. In this context, this thesis addresses the general problem of knowledge representation in Smart Spaces. The main objective is to develop knowledge-based models, equipped with semantics to learn, infer and monitor human behaviours in Smart Spaces. Moreover, it is easy to recognize that some aspects of this problem have a high degree of uncertainty, and therefore, the developed models must be equipped with mechanisms to manage this type of information. A fuzzy ontology and a semantic hybrid system are presented to allow modelling and recognition of a set of complex real-life scenarios where vagueness and uncertainty are inherent to the human nature of the users that perform it. The handling of uncertain, incomplete and vague data (i.e., missing sensor readings and activity execution variations, since human behaviour is non-deterministic) is approached for the rst time through a fuzzy ontology validated on real-time settings within a hybrid data-driven and knowledgebased architecture. The semantics of activities, sub-activities and real-time object interaction are taken into consideration. The proposed framework consists of two main modules: the low-level sub-activity recognizer and the high-level activity recognizer. The rst module detects sub-activities (i.e., actions or basic activities) that take input data directly from a depth sensor (Kinect). The main contribution of this thesis tackles the second component of the hybrid system, which lays on top of the previous one, in a superior level of abstraction, and acquires the input data from the rst module's output, and executes ontological inference to provide users, activities and their in uence in the environment, with semantics. This component is thus knowledge-based, and a fuzzy ontology was designed to model the high-level activities. Since activity recognition requires context-awareness and the ability to discriminate among activities in di erent environments, the semantic framework allows for modelling common-sense knowledge in the form of a rule-based system that supports expressions close to natural language in the form of fuzzy linguistic labels. The framework advantages have been evaluated with a challenging and new public dataset, CAD-120, achieving an accuracy of 90.1% and 91.1% respectively for low and high-level activities. This entails an improvement over both, entirely data-driven approaches, and merely ontology-based approaches. As an added value, for the system to be su ciently simple and exible to be managed by non-expert users, and thus, facilitate the transfer of research to industry, a development framework composed by a programming toolbox, a hybrid crisp and fuzzy architecture, and graphical models to represent and con gure human behaviour in Smart Spaces, were developed in order to provide the framework with more usability in the nal application. As a result, human behaviour recognition can help assisting people with special needs such as in healthcare, independent elderly living, in remote rehabilitation monitoring, industrial process guideline control, and many other cases. This thesis shows use cases in these areas.

A carbohydrate-based mechanism of species recognition in sea urchin fertilization

In the present review, we describe a systematic study of the sulfated polysaccharides from marine invertebrates, which led to the discovery of a carbohydrate-based mechanism of sperm-egg recognition during sea urchin fertilization. We have described unique polymers present in these organisms, especially sulfated fucose-rich compounds found in the egg jelly coat of sea urchins. The polysaccharides have simple, linear structures consisting of repeating units of oligosaccharides. They differ among the various species of sea urchins in specific patterns of sulfation and/or position of the glycosidic linkage within their repeating units. These polysaccharides show species specificity in inducing the acrosome reaction in sea urchin sperm, providing a clear-cut example of a signal transduction event regulated by sulfated polysaccharides. This distinct carbohydrate-mediated mechanism of sperm-egg recognition coexists with the bindin-protein system. Possibly, the genes involved in the biosynthesis of these sulfated fucans did not evolve in concordance with evolutionary distance but underwent a dramatic change near the tip of the Strongylocentrotid tree. Overall, we established a direct causal link between the molecular structure of a sulfated polysaccharide and a cellular physiological event - the induction of the sperm acrosome reaction in sea urchins. Small structural changes modulate an entire system of sperm-egg recognition and species-specific fertilization in sea urchins. We demonstrated that sulfated polysaccharides - in addition to their known function in cell proliferation, development, coagulation, and viral infection - mediate fertilization, and respond to evolutionary mechanisms that lead to species diversity.

Design and implementation of machine vision system for the mobile assembly robot

The review of intelligent machines shows that the demand for new ways of helping people in perception of the real world is becoming higher and higher every year. This thesis provides information about design and implementation of machine vision for mobile assembly robot. The work has been done as a part of LUT project in Laboratory of Intelligent Machines. The aim of this work is to create a working vision system. The qualitative and quantitative research were done to complete this task. In the first part, the author presents the theoretical background of such things as digital camera work principles, wireless transmission basics, creation of live stream, methods used for pattern recognition. Formulas, dependencies and previous research related to the topic are shown. In the second part, the equipment used for the project is described. There is information about the brands, models, capabilities and also requirements needed for implementation. Although, the author gives a description of LabVIEW software, its add-ons and OpenCV which are used in the project. Furthermore, one can find results in further section of considered thesis. They mainly represented by screenshots from cameras, working station and photos of the system. The key result of this thesis is vision system created for the needs of mobile assembly robot. Therefore, it is possible to see graphically what was done on examples. Future research in this field includes optimization of the pattern recognition algorithm. This will give less response time for recognizing objects. Presented by author system can be used also for further activities which include artificial intelligence usage.

Recognition of Facial Expressions with Autoencoders and Convolutional-Nets

Les humains communiquent via différents types de canaux: les mots, la voix, les gestes du corps, des émotions, etc. Pour cette raison, un ordinateur doit percevoir ces divers canaux de communication pour pouvoir interagir intelligemment avec les humains, par exemple en faisant usage de microphones et de webcams. Dans cette thèse, nous nous intéressons à déterminer les émotions humaines à partir d’images ou de vidéo de visages afin d’ensuite utiliser ces informations dans différents domaines d’applications. Ce mémoire débute par une brève introduction à l'apprentissage machine en s’attardant aux modèles et algorithmes que nous avons utilisés tels que les perceptrons multicouches, réseaux de neurones à convolution et autoencodeurs. Elle présente ensuite les résultats de l'application de ces modèles sur plusieurs ensembles de données d'expressions et émotions faciales. Nous nous concentrons sur l'étude des différents types d’autoencodeurs (autoencodeur débruitant, autoencodeur contractant, etc) afin de révéler certaines de leurs limitations, comme la possibilité d'obtenir de la coadaptation entre les filtres ou encore d’obtenir une courbe spectrale trop lisse, et étudions de nouvelles idées pour répondre à ces problèmes. Nous proposons également une nouvelle approche pour surmonter une limite des autoencodeurs traditionnellement entrainés de façon purement non-supervisée, c'est-à-dire sans utiliser aucune connaissance de la tâche que nous voulons finalement résoudre (comme la prévision des étiquettes de classe) en développant un nouveau critère d'apprentissage semi-supervisé qui exploite un faible nombre de données étiquetées en combinaison avec une grande quantité de données non-étiquetées afin d'apprendre une représentation adaptée à la tâche de classification, et d'obtenir une meilleure performance de classification. Finalement, nous décrivons le fonctionnement général de notre système de détection d'émotions et proposons de nouvelles idées pouvant mener à de futurs travaux.