Speaking and presenting in public are critical skills for academic and professional development. These skills are demanded across society, and their development and evaluation are a challenge faced by higher education institutions. There are some challenges to evaluate objectively, as well as to generate valuable information to professors and appropriate feedback to students. In this paper, in order to understand and detect patterns in oral student presentations, we collected data from 222 Computer Engineering (CE) fresh students at three different times, over two different years (2017 and 2018). For each presentation, using a developed system and Microsoft Kinect, we have detected 12 features related to corporal postures and oral speaking. These features were used as input for the clustering and statistical analysis that allowed for identifying three different clusters in the presentations of both years, with stronger patterns in the presentations of the year 2017. A Wilcoxon rank-sum test allowed us to evaluate the evolution of the presentations attributes over each year and pointed out a convergence in terms of the reduction of the number of features statistically different between presentations given at the same course time. The results can further help to give students automatic feedback in terms of their postures and speech throughout the presentations and may serve as baseline information for future comparisons with presentations from students coming from different undergraduate courses.
The present paper describes a clustering based approach to identify the main corporal patterns in students oral presentations during a given course. Data from 43 students presentations was collected through the use of Microsoft Kinect. The 16 collected features were used as input information in the clusterization process allowing the identification of three main profiles of presenters: passive, active, and semi-active. An analysis of the evolution of these profiles during the semester points out a decrease in the percentage of the passive profile throughout the course, and an increase in the percentage of the semi-active profile. These different profiles will be integrated into the system that collects the postures information in order to allow the automated classification of the presenters in real time.Resumo. O presente trabalho apresenta uma abordagem baseada em clusterização para identificar os principais padrões corporais em apresentações orais de estudantes em uma disciplina. Utilizando o Microsoft Kinect, foram coletados dados de 43 apresentações de estudantes realizadas em três momentos distintos ao longo do semestre. As 16 características coletadas pelo sistema desenvolvido foram utilizadas como entrada para a clusterização que permitiu identificar três perfis principais de apresentadores: passivos, ativos e semiativos. Uma análise sobre a evolução desses perfis aponta que houve uma diminuição do percentual do perfil passivo ao longo do semestre e um aumento do percentual do perfil semi-ativo. Esses tipos de perfis serão integrados ao sistema de coleta de posturas para a futura classificação automática dos apresentadores em tempo real.
The quality control is an essential step in fabric industries. Detectdefects in the early stages can reduce costs and increase the qualityof the products. Currently, this task is mainly done by humans,whose judgment can be affected by fatigue. Computer vision-basedtechniques can automatically detect defects, reducing the need forhuman intervention. In this context, this work proposes an imageblock-processing approach, where we compare the Segmentation-Based Fractal Texture Analysis, Gray Level Co-Occurrence Matrix,and Local Binary Pattern in the feature extraction step. Aimingto show the efficiency of this approach for the problem, these resultswere compared with the same algorithms without the blockprocessingapproach. A Support Vector Machine optimized by Grid-Search Algorithm was used to classify the fabrics. The databaseused, which is available online, is composed of 479 images fromsamples with defects and without it. The results show that thisblock processing approach can improve the classification results,achieving 100% in this work.
The present paper describes a sequential pattern mining based approach to identify the main corporal sequences in students oral presentations during a given course. Data from students presentations was collected through the use of Microsoft Kinect and Leikelen software, the total number of observations was 65. The 7 collected features were used as input information in the SPMF tool, allowing the identification of main sequences of presenters. Sequences with the Hands Down attribute were the most frequent in all presentations. It has also been found that the presentations 1 and 3 are more similar in terms of sequence than with the second. The evaluation of the sequences can be integrated into the tool so that the teacher can return feedback to the students about their postures.Resumo. O presente trabalho descreve uma abordagem de Mineração de Padrão Sequencial para identificar as principais sequências corporais em apresentações orais de estudantes durante um determinado curso. Os dados das apresentações dos alunos foram coletados através do uso do Microsoft Kinect e do software Leikelen, totalizando 65 observações. As 7 características coletadas foram utilizadas como informações de entrada na ferramenta SPMF, permitindo a identificação das principais sequências dos apresentadores. Sequências com o atributo Mãos Baixas foram as mais frequentes em todas as apresentações. Verificou-se também que as apresentações 1 e 3 são mais semelhantes em termos de sequência do que com a segunda. A avaliação das sequências pode ser integrada na ferramenta para que o professor possa retornar feedback aos alunos sobre suas posturas.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.