This thesis provides insights into how technology can be and is used as child-centric learning tools within primary school classrooms. The conducted studies look closely at how tablet technology is integrated into the modern classroom, and considers how existing digital writing technologies could support handwriting-based learning exercises in future. This is achieved by conducting three in-the-wild studies, using different approaches, with a total of seventy-four children in school classrooms. In the first study, focus is placed on how tablets integrate into and with existing classroom practices, documenting when and how children use tablets in class. Relevant and complementary to this, the use of traditional writing tools is questioned and two further studies explore the potential and suitability of digital pens to support children's handwriting-based learning. One looks in detail at how children's handwriting is effected by different existing digital pen technologies. The other study, conducted through a creative, participatory design session, asks children to provide their opinions regarding desirable features for digital writing technology. The findings from this research classify and exemplify the role of tablets in the classroom, and explore potential design directions of digital writing tools which could be used by children in the future. This work may be useful and of interest to others who conduct research with children within the fields of Human Computer Interaction, Child Computer Interaction or education.

The aim of this research is to investigate how handwriting as an individualisation technique in fraud investigation can be used by police detectives, SARS investigators and forensic investigation specialists, who are responsible for the investigation and linking the perpetrator, with a view to criminal prosecution. A further intent was to share and introduce a number of important concepts, namely: criminal investigation, identification, individualisation, fraud, evidence and handwriting. The research will explain the sophisticated investigation techniques used to obtain sufficient information to prove the true facts in a court of law. Identification is the collective aspect of the set of characteristics by which an object is definitively recognisable or known, while the individual characteristics establish the individuality of a specific object. Many types of evidence may be used to link an individual with a crime scene, and associate that individual with the performed illegal handling. It also explained that during a cheque/document fraud investigation, it is in most cases the only link to information to trace, identify and individualise the perpetrator, and to obtain a handwriting specimen. It is also discussed how to eliminate a person being a writer of a document, and how to collect, package and mark a disputed document during the investigation. If the investigators use their knowledge of these concepts, it should enhance their investigative skills, and empower them to be become better equipped for the challenges they face in identifying, individualising and linking the perpetrators, in order to ensure successful prosecution and conviction. / Police Practice / M.Tech. (Forensic Investigation)

Atualmente, a temática da dislexia e do transtorno déficit de atenção/hiperatividade passou a ocupar os mais diversos espaços acadêmicos e políticos, com manchetes de especialistas nos programas de televisão, rádio, jornais, e criação de diversos projetos de lei que se propõem a criar serviços de diagnóstico e tratamento nas secretarias de educação. Em todos esses espaços sociais, estes supostos distúrbios são apresentados como doenças neurológicas, que explicariam dificuldades encontradas pelas crianças em seu processo de escolarização. Se, por um lado, temos este quadro de afirmação da suposta doença; por outro há um conjunto de autores que têm questionado tais distúrbios e reiterado a necessidade de compreender a complexidade do processo de alfabetização das crianças iniciantes (no caso da dislexia) e todo o contexto sociocultural que envolve o comportamento das crianças na atualidade (no caso do TDAH). O presente trabalho, por meio de uma abordagem qualitativa de estudo de caso, investiga a história do processo de escolarização, a produção do diagnóstico de dislexia e seus efeitos nas relações escolares de crianças em fase inicial de aquisição da leitura e da escrita. Foram realizadas entrevistas com a mãe, coordenadora pedagógica, professoras e criança diagnosticada, além da análise do laudo realizado por equipe multidisciplinar. No discurso da mãe sobre a história escolar do filho surge o tema das dificuldades escolares trazido como um problema da criança, que teria algo a menos ou em quem faltaria algo a mais. Esta concepção instaura um processo diagnóstico, gerando um tratamento que constitui o processo de medicalização e de culpabilização da criança e de sua família pelo não aprender na escola. As vozes das professoras, não escutadas durante o processo diagnóstico, denunciam que diferentes concepções de desenvolvimento, de aprendizagem e crenças sobre os alunos resultam em relações, ações pedagógicas e, portanto, possibilidades de aprendizagem também distintas. O diagnóstico encontrado, foi realizado ao largo da escola o que evidencia a concepção de desenvolvimento humano na qual se pauta, partindo do pressuposto que a dificuldade pertence à criança. A avaliação incluiu apenas aplicação de testes de diversas áreas tais como psicologia, fonoaudiologia e neurologia, ignorando resultados de pesquisas recentes que inviabilizam o uso de alguns deles por não estarem relacionados ao alegado distúrbio e utilizando outros relacionados exatamente ao motivo do encaminhamento para a avaliação: questões de leitura e escrita. Ter um diagnóstico de dislexia cristaliza um movimento, um processo dinâmico que é o de aprendizagem e desenvolvimento. Desta forma o diagnóstico segue orientado somente para a falta e para as dificuldades estabelecendo limites a priori para o desenvolvimento do sujeito. Além desses efeitos relacionados à aprendizagem, existem outros decorrentes da medicação que parecem inerentes ao diagnóstico de dislexia acompanhado de TDAH. A criança vive na escola relações estigmatizadas que contribuíram na constituição de sua subjetividade, pautada na doença e nas limitações impostas pelo rótulo diagnóstico / Nowadays dyslexia and ADDH (Attention Deficit Disorder with Hyperactivity) topics are gaining larger place in the media. This stimulates a series of law projects intended to create public services of diagnosis and treatment. In this attempt of controlling and managing the problem, these so called \"disorders\", are rather quickly defined by operators and specialists as neurological diseases, which would logically explain the difficulties incurred by children during their schooling process and partially solve teachers and parents headaches on that matter. If practitioners now tend to legitimate this supposed illness, on the other hand, there are many serious authors questioning such straight definition, empathizing two ideas: the urgent necessity of fully understand the complexity of literacy process in beginner children and the necessity to grasp the sociocultural context that surrounds children education and behavior in the present time. This work investigates, by means of a qualitative case study, the diagnosis dyslexia of a child during the schooling process acquiring reading and writing abilities, and consequences of this diagnosis on his relations at school. We hereby present interviews with the mother, the pedagogical coordinator, with the teachers and with the diagnosed child himself. The study also presents the analysis of the report made by the multidisciplinary team responsible for evaluation. The mother of the child that is object of the study presents her sons problems as a lack of something that differentiates him from others in the educational process. This conception establishes the base for the diagnosis process, and results in a treatment that starts medicalization, blaming the child and his family and justifying the inability to learn at school. The voice of the teachers, unheard during the diagnosis process, denounces that different conceptions of development and learning as well as different beliefs about the students end up in distinguished relations, pedagogical actions and learning possibilities for the pupils. The diagnosis found was achieved without considering the school, which highlights the idea of human development on the basis of this approach, assuming as its starting point that the difficulty belongs to the child. The evaluation included tests in many areas such as psychology, phonoaudiology and neurology; ignoring the results of recent researches that discredit some of these methods for not being related with dyslexia. They also use other methods, which are connected exactly to the motifs why the evaluation was requested at the first place: questions referring to reading and writing ability. Declaring dyslexia crystallizes a movement, a dynamic process that is learning and developing. By this way the diagnosis focuses on the difficulties building up barriers to growth of the subject. Besides those effects related to learning, there are other effects, due to medication, that seems inherent to the diagnosis of dyslexia followed by the one of ADDH. The child lives stigmatized relations at school that contribute to the constitution of his subjectivity, based on the illness and limitations imposed by the diagnostic label

O reconhecimento de padrões por computador é uma das mais importantes ferramentas da Inteligência Artificial presente em inúmeras áreas do conhecimento com aplicações em diversos setores, incluindo o reconhecimento de caracteres. O objetivo da dissertação se concentra na investigação de um processo computacional automatizado - Sistema Computacional Paraconsistente - capaz de reconhecer Caracteres Numéricos Manuscritos e Caracteres Magnéticos Codificados em 7 Barras utilizados em cheques bancários brasileiros, fornecendo uma fundamentação técnica para reconhecer documentos e imagens digitalizadas e, também, sinais biológicos. Embora haja vários estudos em reconhecimento de caracteres, optou-se pelo estudo desse tema devido à sua intrínseca importância e constante desenvolvimento, além de possibilitar adaptações para fazer o reconhecimento de diferentes tipos de sinais como, por exemplo, sinais biológicos. A metodologia adotada para essa tarefa se baseia nas Redes Neurais Artificiais Paraconsistentes por se tratar de uma ferramenta com capacidade de trabalhar com dados imprecisos, inconsistentes e paracompletos sem o perigo de trivialização. O processo de reconhecimento desse sistema é realizado a partir de algumas características do caractere previamente selecionadas com base em algumas técnicas do Grafismo e realiza-se a análise dessas características bem como o reconhecimento do caractere através das Redes Neurais Artificiais Paraconsistentes O sistema foi construído para reconhecer caracteres numéricos com um padrão previamente definido, onde adotou-se os Caracteres Magnéticos Codificados em 7 Barras utilizados em cheques bancários e, posteriormente, o sistema foi aperfeiçoado para fazer o reconhecimento de Caracteres Numéricos Manuscritos. Para a validação do estudo proposto apresentou-se dados reais, a saber, lotes de cheques e caracteres numéricos manuscritos digitalizados onde o sistema apresentou 97,85% de acertos para os Caracteres Magnéticos Codificados em 7 Barras e 91,62% de acertos para Caracteres Numéricos Manuscritos. O resultado obtido demonstra que o sistema é robusto o suficiente e pode servir de estudo para a análise de sinais em áreas correlatas com nível de precisão semelhante / Computer pattern recognition is one of the most important Artificial Intelligence tools present in numerous knowledge areas with applications in several themes, including the character recognition. The aim of this dissertation is the investigation of an automated computational process - Paraconsistent Computational System - able to recognize Handwritten Numeric Characters and Magnetic Ink Character Recognition used on Brazilian bank checks furnishing a technical basis to recognize digital documents, digital images and biological signals. Although there are several studies on character recognition, it was chosen to study this theme due to its intrinsic importance and constant improvement, besides enabling adjustments to the recognition of different kinds of signals such as, biological signals. The methodology employed for the task is based on Paraconsistent Artificial Neural Networks for being a tool with the ability to work with imprecise, inconsistent and paracomplete data without trivialization. The recognition process of this system is performed from some previously selected character features based on some Graphics techniques and, it performs the analysis of these features as well as the character recognition are performed through the Paraconsistent Artificial Neural Networks. The system was built to recognize numeric characters with a previously defined pattern where it was chosen the Magnetic Ink Character Recognition used on Brazilian bank checks and then the system was improved to recognize handwritten numeric characters. It was presented real data as checks\' batches and scanned handwritten numeric characters to validate the proposed study and the system reached 97.85% hits for Magnetic Ink Character Recognition and 91.62% hits for Handwritten Numeric Characters. The obtained result demonstrates that the system is robust enough for signal analysis study in correlated areas with similar precision level

Avec l’avènement des dispositifs nomades tels que les smartphones et les tablettes, la reconnaissance automatique de l’écriture manuscrite cursive à partir d’un signal en ligne est devenue durant les dernières décennies un besoin réel de la vie quotidienne à l’ère numérique. Dans le cadre de cette thèse, nous proposons de nouvelles stratégies pour un système de reconnaissance de mots manuscrits en-ligne. Ce système se base sur une méthode collaborative segmentation/reconnaissance et en utilisant des analyses à deux niveaux : caractère et bi-caractères. Plus précisément, notre système repose sur une segmentation de mots manuscrits en graphèmes afin de créer un treillis à L niveaux. Chaque noeud de ce treillis est considéré comme un caractère potentiel envoyé à un moteur de Reconnaissance de Caractères Isolés (RCI) basé sur un SVM. Pour chaque noeud, ce dernier renvoie une liste de caractères associés à une liste d’estimations de probabilités de reconnaissance. Du fait de la grande diversité des informations résultant de la segmentation en graphèmes, en particulier à cause de la présence de morceaux de caractères et de ligatures, l’injection de chacun des noeuds du treillis dans le RCI engendre de potentielles ambiguïtés au niveau du caractère. Nous proposons de lever ces ambiguïtés en utilisant des modèles de bi-caractères, basés sur une régression logistique dont l’objectif est de vérifier la cohérence des informations à un niveau de reconnaissance plus élevé. Finalement, les résultats renvoyés par le RCI et l’analyse des modèles de bi-caractères sont utilisés dans la phase de décodage pour parcourir le treillis dans le but de trouver le chemin optimal associé à chaque mot dans le lexique. Deux méthodes de décodage sont proposées (recherche heuristique et programmation dynamique), la plus efficace étant basée sur de la programmation dynamique. / With the advent of mobile devices such as tablets and smartphones over the last decades, on-line handwriting recognition has become a very highly demanded service for daily life activities and professional applications. This thesis presents a new approach for on-line handwriting recognition. This approach is based on explicit segmentation/recognition integrated in a two level analysis system: character and bi-character. More specifically, our system segments a handwritten word in a sequence of graphemes to be then used to create a L-levels lattice of graphemes. Each node of the lattice is considered as a character to be submitted to a SVM based Isolated Character Recognizer (ICR). The ICR returns a list of potential character candidates, each of which is associated with an estimated recognition probability. However, each node of the lattice is a combination of various segmented graphemes. As a consequence, a node may contain some ambiguous information that cannot be handled by the ICR at character level analysis. We propose to solve this problem using "bi-character" models based on Logistic Regression, in order to verify the consistency of the information at a higher level of analysis. Finally, the recognition results provided by the ICR and the bi-character models are used in the word decoding stage, whose role is to find the optimal path in the lattice associated to each word in the lexicon. Two methods are presented for word decoding (heuristic search and dynamic programming), and dynamic programming is found to be the most effective.

Nous présentons dans cette thèse un nouveau paradigme pour caractériser la maladie d’Alzheimer à travers l’écriture manuscrite acquise sur tablette graphique. L’état de l’art est dominé par des méthodes qui supposent un comportement unique ou homogène au sein de chaque profil cognitif. Ces travaux exploitent des paramètres cinématiques globaux, sur lesquels ils appliquent des tests statistiques ou des algorithmes de classification pour discriminer les différents profils cognitifs (les patients Alzheimer, les troubles cognitifs légers (« Mild Cognitive impairment » : MCI) et les sujets Contrôle (HC)). Notre travail aborde ces deux limites de la littérature de la façon suivante : premièrement au lieu de considérer un comportement homogène au sein de chaque profil cognitif ou classe (HC, MCI, ES-AD : « Early-Stage Alzheimer Disease »), nous nous sommes affranchis de cette hypothèse (ou contrainte) forte de la littérature. Nous considérons qu’il peut y avoir plusieurs comportements au sein de chaque profil cognitif. Ainsi, nous proposons un apprentissage semi-supervisé pour trouver des groupes homogènes de sujets et analysons l’information contenue dans ces clusters ou groupes sur les profils cognitifs. Deuxièmement, au lieu d’exploiter les paramètres cinématiques globaux (ex : vitesse moyenne, pression moyenne, etc.), nous avons défini deux paramétrisations ou codages : une paramétrisation semi-globale, puis locale en modélisant la dynamique complète de chaque paramètre. L’un de nos résultats importants met en évidence deux clusters majeurs qui sont découverts, l’un dominé par les sujets HC et MCI et l’autre par les MCI et ES-AD, révélant ainsi que les patients atteints de MCI ont une motricité fine qui est proche soit des sujets HC, soit des patients ES-AD. Notre travail montre également que la vitesse prise localement regroupe un ensemble riche des caractéristiques telles que la taille, l’inclinaison, la fluidité et la régularité, et révèle comment ces paramètres spatiotemporels peuvent conjointement caractériser les profils cognitifs. / We present, in this thesis, a novel paradigm for assessing Alzheimer’s disease by analyzing impairment of handwriting (HW) on tablets, a challenging problem that is still in its infancy. The state of the art is dominated by methods that assume a unique behavioral trend for each cognitive profile, and that extract global kinematic parameters, assessed by standard statistical tests or classification models, for discriminating the neuropathological disorders (Alzheimer’s (AD), Mild Cognitive Impairment (MCI)) from Healthy Controls (HC). Our work tackles these two major limitations as follows. First, instead of considering a unique behavioral pattern for each cognitive profile, we relax this heavy constraint by allowing the emergence of multimodal behavioral patterns. We achieve this by performing semi-supervised learning to uncover homogeneous clusters of subjects, and then we analyze how much information these clusters carry on the cognitive profiles. Second, instead of relying on global kinematic parameters, mostly consisting of their average, we refine the encoding either by a semi-global parameterization, or by modeling the full dynamics of each parameter, harnessing thereby the rich temporal information inherently characterizing online HW. Thanks to our modeling, we obtain new findings that are the first of their kind on this research field. A striking finding is revealed: two major clusters are unveiled, one dominated by HC and MCI subjects, and one by MCI and ES-AD, thus revealing that MCI patients have fine motor skills leaning towards either HC’s or ES-AD’s. This thesis introduces also a new finding from HW trajectories that uncovers a rich set of features simultaneously like the full velocity profile, size and slant, fluidity, and shakiness, and reveals, in a naturally explainable way, how these HW features conjointly characterize, with fine and subtle details, the cognitive profiles.

We use a Fully Convolutional Neural Network (FCNN) to classify pixels in historical document images, enabling the extraction of high-quality, pixel-precise and semantically consistent layers of masked content. We also analyze a dataset of hand-labeled historical form images of unprecedented detail and complexity. The semantic categories we consider in this new dataset include handwriting, machine-printed text, dotted and solid lines, and stamps. Segmentation of document images into distinct layers allows handwriting, machine print, and other content to be processed and recognized discriminatively, and therefore more intelligently than might be possible with content-unaware methods. We show that an efficient FCNN with relatively few parameters can accurately segment documents having similar textural content when trained on a single representative pixel-labeled document image, even when layouts differ significantly. In contrast to the overwhelming majority of existing semantic segmentation approaches, we allow multiple labels to be predicted per pixel location, which allows for direct prediction and reconstruction of overlapped content. We perform an analysis of prevalent pixel-wise performance measures, and show that several popular performance measures can be manipulated adversarially, yielding arbitrarily high measures based on the type of bias used to generate the ground-truth. We propose a solution to the gaming problem by comparing absolute performance to an estimated human level of performance. We also present results on a recent international competition requiring the automatic annotation of billions of pixels, in which our method took first place.

We use a Fully Convolutional Neural Network (FCNN) to classify pixels in historical document images, enabling the extraction of high-quality, pixel-precise and semantically consistent layers of masked content. We also analyze a dataset of hand-labeled historical form images of unprecedented detail and complexity. The semantic categories we consider in this new dataset include handwriting, machine-printed text, dotted and solid lines, and stamps. Segmentation of document images into distinct layers allows handwriting, machine print, and other content to be processed and recognized discriminatively, and therefore more intelligently than might be possible with content-unaware methods. We show that an efficient FCNN with relatively few parameters can accurately segment documents having similar textural content when trained on a single representative pixel-labeled document image, even when layouts differ significantly. In contrast to the overwhelming majority of existing semantic segmentation approaches, we allow multiple labels to be predicted per pixel location, which allows for direct prediction and reconstruction of overlapped content. We perform an analysis of prevalent pixel-wise performance measures, and show that several popular performance measures can be manipulated adversarially, yielding arbitrarily high measures based on the type of bias used to generate the ground-truth. We propose a solution to the gaming problem by comparing absolute performance to an estimated human level of performance. We also present results on a recent international competition requiring the automatic annotation of billions of pixels, in which our method took first place.

This paper concerns methods of verification of person's signature and handwriting. Some of commonly used techniques are resumed and described with related literature being referred. Next aim of this work is design and implementation of a simple handwriting verification application. Application is based on edge detection and comparison of a set of structural and statistical features. As a support classification tool a SVM classifier of the LIBSVM software is employed. The Application is written in C language using OpenCV graphics library. Testing and training set was extracted from samples found in the IAM Handwriting Database. Application was created and tested in the Windows XP operating system.

Les méthodes à l’état de l’art de la reconnaissance de l’écriture sont fondées sur des réseaux de neurones récurrents (RNN) à cellules LSTM ayant des performances remarquables. Dans cette thèse, nous proposons deux nouveaux principes la vérification lexicale et la génération de cohorte afin d’attaquer les problèmes de la reconnaissance de l’écriture : i) le problème des grands lexiques et des décodages dirigés par le lexique ii) la problématique de combinaison de modèles optiques pour une meilleure reconnaissance iii) la nécessité de constituer de très grands ensembles de données étiquetées dans un contexte d’apprentissage profond. La vérification lexicale est une alternative aux décodages dirigés par le lexique peu étudiée à cause des faibles performances des modèles optiques historiques (HMM). Nous montrons dans cette thèse qu’elle constitue une alternative intéressante aux approches dirigées par le lexique lorsqu’elles s’appuient sur des modèles optiques très performants comme les RNN LSTM. La génération de cohorte permet de générer facilement et rapidement un grand nombre de réseaux récurrents complémentaires en un seul apprentissage. De ces deux techniques nous construisons et proposons un nouveau schéma de cascade pour la reconnaissance de mots isolés, une nouvelle combinaison au niveau ligne LV-ROVER et une nouvelle stratégie d’auto-apprentissage de RNN LSTM pour la reconnaissance de mots isolés. La cascade proposée permet de combiner avec la vérification lexicale des milliers de réseaux et atteint des résultats à l’état de l’art pour les bases Rimes et IAM. LV-ROVER a une complexité réduite par rapport à l’algorithme original ROVER et permet de combiner des centaines de réseaux sans modèle de langage tout en dépassant l’état de l’art pour la reconnaissance de lignes sur le jeu de donnéesRimes. Notre stratégie d’auto-apprentissage permet d’apprendre à partir d’un seul réseau BLSTM et sans paramètres grâce à la cohorte et la vérification lexicale, elle montre d’excellents résultats sur les bases Rimes et IAM. / State-of-the-art methods for handwriting recognition are based on LSTM recurrent neural networks (RNN) which achieve high performance recognition. In this thesis, we propose the lexicon verification and the cohort generation as two new building blocs to tackle the problem of handwriting recognition which are : i) the large vocabulary problem and the use of lexicon driven methods ii) the combination of multiple optical models iii) the need for large labeled dataset for training RNN. The lexicon verification is an alternative to the lexicon driven decoding process and can deal with lexicons of 3 millions words. The cohort generation is a method to get easily and quickly a large number of complementary recurrent neural networks extracted from a single training. From these two new techniques we build and propose a new cascade scheme for isolated word recognition, a new line level combination LV-ROVER and a new self-training strategy to train LSTM RNN for isolated handwritten words recognition. The proposed cascade combines thousands of LSTM RNN with lexicon verification and achieves state-of-the art word recognition performance on the Rimes and IAM datasets. The Lexicon Verified ROVER : LV-ROVER, has a reduce complexity compare to the original ROVER algorithm and combine hundreds of recognizers without language models while achieving state of the art for handwritten line text on the RIMES dataset. Our self-training strategy use both labeled and unlabeled data with the unlabeled data being self-labeled by its own lexicon verified predictions. The strategy enables self-training with a single BLSTM and show excellent results on the Rimes and Iam datasets.