Doutorado em Ciência da Computação

URI Permanente para esta coleção

http://repositorio.ufes.br/handle/10/17626

Nível: Doutorado
Ano de início:
Conceito atual na CAPES:
Ato normativo:
Periodicidade de seleção:
Área(s) de concentração:
Url do curso:

Navegar

Navegando Doutorado em Ciência da Computação por Autor "Baduê, Claudine Santos"

Agora exibindo 1 - 2 de 2
  • Nenhuma Miniatura disponível
    Item
    Exames inteligentes: evidenciação estatística de perfis de aprendizagem, composição de banco de itens multidimensionais e personalização de avaliação
    (Universidade Federal do Espírito Santo, 2023-05-19) Silva, Wesley Pereira da; Oliveira, Elias Silva de; https://orcid.org/0000-0003-2066-7980; http://lattes.cnpq.br/2210356035827181; http://lattes.cnpq.br/8881034997521890; Baduê, Claudine Santos; https://orcid.org/0000-0003-1810-8581; http://lattes.cnpq.br/1359531672303446; Azevedo, Caio Lucidius Naberezny; https://orcid.org/0000-0001-9535-292X; http://lattes.cnpq.br/0856524274837137; Santos, Thiago Oliveira dos; https://orcid.org/0000-0001-7607-635X; http://lattes.cnpq.br/5117339495064254; Guzman, Jorge Luis Bazan
    A common challenge to the areas of knowledge is the construction of teaching strategies that are sufficiently general to suit audiences with varied learning profiles. Usually, the teacher defines his teaching plan according to learning objectives, which are evaluated through the expression of latent traits that denote proficiency on the part of the subject being examined. Thus, intelligent techniques and technological tools are opportune to contribute to increasing the quality of teaching and reducing the teaching effort in the execution of complex activities such as, for example: formulation of assessment items, application of tests and provision of feedback to students. The formal rigor in the creation of instruments for assessment, tabulation and calculation of grades is a key factor to avoid bias in conducting the assessment of learning and estimating the ability of students. Student performance is the first dimension to be considered in the assessment. The grouping of similar performances allows characterizing groups that represent learning profiles. Self-assessment and peer assessment are techniques to stimulate the student’s self-criticism in relation to himself and his classmates, seeking to discourage evaluative biases by encouraging the examinee’s coherence when exercising the role of evaluator. The logistic models derived from Psychometrics allow the quantitative characterization of the evaluation items, allowing the measurement of qualitative aspects, such as: difficulty, discrimination and propensity to kick. With psychometric models, the probability of success of the subject can be predicted when being evaluated with a certain item. Finally, the use of Natural Language Processing provides the selection of items by content similarity with a search expression, which represents a subject to be retrieved in a set of the test items bank. In this way, we seek to propose a method of creating individualized assessment trail, composed of a sequence of activities in a certain order appropriate to the ability of the examinee. Thus, we present an intelligent computerized adaptive test approach, whose execution configuration is adjustable to qualitative, quantitative and/or content teaching strategies related to pre-defined terms. The contribution envisaged with such a proposal is to extrapolate a two-dimensional parameter space of the evaluations, composed of the examinee’s performances and scores achieved by item; for a multidimensional space that considers the characteristics of the items in psychometric and semantic terms, as well as the characteristics of the examinees and historical data of subjects with similar trajectories.
  • Nenhuma Miniatura disponível
    Item
    Generating road grid maps and path plans for self-driving cars using laser remission data and deep neural networks
    (Universidade Federal do Espírito Santo, 2024-09-24) Carneiro, Raphael Vivacqua; Souza, Alberto Ferreira de; https://orcid.org/0000-0003-1561-8447; Baduê, Claudine Santos; Rauber, Thomas Walter; Komati, Karin Satie; Andrade, Mariella Berger
    This work proposes the use of deep neural networks (DNN) for solving the problem of inferring the location of drivable lanes of roadways and their relevant properties such as the lane change right-of-way, even if the line markings are poor or absent. This problem is relevant to the operation of self-driving cars which requires precise maps and precise path plans. Our approach to the problem is the use of a DNN for semantic segmentation of LiDAR remission grid maps into road grid maps. Both LiDAR remission grid maps and road grid maps are square matrices in which each cell represents features of a small 2D-squared region of the real world (e.g., 20cm × 20cm). A LiDAR remission grid map cell contains the information about the average intensity of laser reflection remission on the surface of that particular place. A road grid map cell contains the semantic information about whether it belongs to either a drivable lane or a line marking or a non-drivable area. The semantic codes associated with the road map cells contain all information required for building a network of valid paths, which are required for self-driving cars to build their path plans. Our proposal is a novel technique for the automatic building of viable path plans for self-driving cars. In our experiments we use the self-driving car of UFES, IARA (Intelligent Autonomous Robotic Automobile). We built datasets of manually marked road lanes and use them to train and validate the DNNs used for the semantic segmentation and the generation of road grid maps from laser remission grid maps. The results achieved an average segmentation accuracy of 94.7% in cases of interest. The path plans automatically generated from the inferred road grid maps were tested in the real world using IARA and has shown performance equivalent to that of manually generated path plans.