Interpretation and explanation of computer vision classification of carambola (Averrhoa carambola L.) according to maturity stage.

The classification of carambola, also known as starfruit, according to quality parameters is usually conducted by trained human evaluators through visual inspections. This is a costly and subjective method that can generate high variability in results. As an alternative, computer vision systems (CVS) combined with deep learning (DCVS) techniques have been introduced in the industry as a powerful and an innovative tool for the rapid and non-invasive classification of fruits. However, validating the learning capability and trustworthiness of a DL model, aka black box, to obtain insights can be challenging. To reduce this gap, we propose an integrated eXplainable Artificial Intelligence (XAI) method for the classification of carambolas at different maturity stages. We compared two Residual Neural Networks (ResNet) and Visual Transformers (ViT) to identify the image regions that are enhanced by a Random Forest (RF) model, with the aim of providing more detailed information at the feature level for classifying the maturity stage. Changes in fruit colour and physicochemical data throughout the maturity stages were analysed, and the influence of these parameters on the maturity stages was evaluated using the Gradient-weighted Class Activation Mapping (Grad-CAM), the Attention Maps using RF importance. The proposed approach provides a visualization and description of the most important regions that led to the model decision, in wide visualization follows the models an importance features from RF. Our approach has promising potential for standardized and rapid carambolas classification, achieving 91 % accuracy with ResNet and 95 % with ViT, with potential application for other fruits.

How people interact with a chatbot against disinformation and fake news in COVID-19 in Brazil: The CoronaAI case.

BACKGROUND: The search for valid information was one of the main challenges encountered during the COVID-19 pandemic, which resulted in the development of several online alternatives. OBJECTIVES: To describe the development of a computational solution to interact with users of different levels of digital literacy on topics related to COVID-19 and to map the correlations between user behavior and events and news that occurred throughout the pandemic. METHOD: CoronaAI, a chatbot based on Google's Dialogflow technology, was developed at a public university in Brazil and made available on WhatsApp. The dataset with users' interactions with the chatbot comprises approximately 7,000 hits recorded throughout eleven months of CoronaAI usage. RESULTS: CoronaAI was widely accessed by users in search of valuable and updated information on COVID-19, including checking the veracity of possible fake news about the spread of cases, deaths, symptoms, tests and protocols, among others. The mapping of users' behavior revealed that as the number of cases and deaths increased and as COVID-19 became closer, users showed a greater need for information applicable to self-care compared to following the statistical data. In addition, they showed that the constant updating of this technology may contribute to public health by enhancing general information on the pandemic and at the individual level by clarifying specific doubts about COVID-19. CONCLUSION: Our findings reinforce the potential usefulness of chatbot technology to resolve a wide spectrum of citizens' doubts about COVID-19, acting as a cost-effective tool against the parallel pandemic of misinformation and fake news.

Football player dominant region determined by a novel model based on instantaneous kinematics variables.

Dominant regions are defined as regions of the pitch where a player can reach before any other and are commonly determined without considering the free-spaces in the pitch. We presented an approach to football players' dominant regions analysis, based on movement models created from players' positions, displacement, velocity, and acceleration vectors. 109 Brazilian male professional football players were analysed during official matches, computing over 15 million positional data obtained by video-based tracking system. Movement models were created based on players' instantaneous vectorial kinematics variables, then probabilities models and dominant regions were determined. Accuracy in determining dominant regions by the proposed model was tested for different time-lag windows. We calculated the areas of dominant, free-spaces, and Voronoi regions. Mean correct predictions of dominant region were 96.56%, 88.64%, and 72.31% for one, two, and three seconds, respectively. Dominant regions areas were lower than the ones computed by Voronoi, with median values of 73 and 171 m2, respectively. A median value of 5537 m2 was presented for free-space regions, representing a large part of the pitch. The proposed movement model proved to be more realistic, representing the match dynamics and can be a useful method to evaluate the players' tactical behaviours during matches.

IoTDS: A One-Class Classification Approach to Detect Botnets in Internet of Things Devices.

Internet of Things (IoT) devices have become increasingly widespread. Despite their potential of improving multiple application domains, these devices have poor security, which can be explored by attackers to build large-scale botnets. In this work, we propose a host-based approach to detect botnets in IoT devices, named IoTDS (Internet of Things Detection System). It relies on one-class classifiers, which model only the legitimate device behaviour for further detection of deviations, avoiding the manual labelling process. The proposed solution is underpinned by a novel agent-manager architecture based on HTTPS, which prevents the IoT device from being overloaded by the training activities. To analyse the device's behaviour, the approach extracts features from the device's CPU utilisation and temperature, memory consumption, and number of running tasks, meaning that it does not make use of network traffic data. To test our approach, we used an experimental IoT setup containing a device compromised by bot malware. Multiple scenarios were made, including three different IoT device profiles and seven botnets. Four one-class algorithms (Elliptic Envelope, Isolation Forest, Local Outlier Factor, and One-class Support Vector Machine) were evaluated. The results show the proposed system has a good predictive performance for different botnets, achieving a mean F1-score of 94% for the best performing algorithm, the Local Outlier Factor. The system also presented a low impact on the device's energy consumption, and CPU and memory utilisation.

Modelagem de banco de dados para diagnóstico de transtornos mentais pela voz / Database modeling for diagnosis mental disorders by voice / Modelado de base de datos para el diagnóstico de los trastornos mentales de voz

Desenvolver um modelo para o banco de dados de vozes brasileiro para armazenar dados do paciente e amostras de áudio capturados, mantendo um histórico de avaliação e tratamento, e possibilitar o diagnóstico automatizado de doenças mentais. Métodos: Pesquisar sobre testes relacionados ao diagnóstico tradicional e pela voz das doenças mentais com a finalidade de promover a abstração de entidades e atributos necessários na constituição da base de dados. Criação do Modelo Entidade-Relacionamento, mapeamento para o Modelo Relacional e criação do Diagrama de Classes. Resultados: Criação do modelo para o Banco de Dados de Vozes Brasileiro. Conclusão: A modelagem foi desenvolvida, de forma que a base de dados possa ser construída e utilizada...

Develop a model for the database of Brazilian voices to store data of patients and captured voices samples. This model supports the automated diagnosis of mental disorders. Method: Research about tests related to tradicional diagnosis and based on voice of mental deseases in order to find the entities and attributes required for database. Creation of the entity-relationship model, mapping to a relational model and the class diagram. Results: Creation the model for the database of Brazilian voices. Conclusion: The modeling was developed and the database can be built and performed...

Desarrollar un modelo para la base de datos de voces brasileñas para almacenar los datos del paciente y muestras de audio captadas por mantener una historia de la evaluación y el tratamiento, y poder hacer un diagnóstico automatizado de la enfermedad mental. Métodos: Se realizó una búsqueda en los diagnósticos tradicionales conexos y la voz de lo trastorno mental con el fin de promover la captación de entidades y atributos requeridos en la constitución de las pruebas de base de datos. Creación del modelo Entidad-Relación, Modelo Relacional para el mapeo y la creación del diagrama de clases. Resultados: La creación del modelo de base de datos para las voces brasileñas. Conclusión: El modelo fue desarrollado para que la base de datos se puede construir y usada...

Reconhecimento automático de Distonia Laríngea com base na sustentação do Pitch / Automatic Laryngeal Dystonia recognition based on Pitch sustainment / Reconocimiento automático de distonía laríngea basado en el apoyo de la Pitch

Objetivos: Desenvolver uma ferramenta automatizada para reconhecimento dos segmentos sem presença de voz durante a fonação do paciente com base na sustentação do Pitch. Método: Os procedimentos para construção e verificação da técnica são a aquisição da voz, janelamento, aplicação da Transformada Discreta de Fourier, detecção do Pitch e a verificação do Pitch. Resultados: Com a análise das 101 vozes, a ferramenta diagnosticou 56 vozes com Distonia Laríngea e 45 como saudáveis. Já o especialista diagnosticou 53 vozes com Distonia Laríngea e 48 como vozes saudáveis. Conclusão: Os resultados demonstraram que o diagnóstico realizado pela ferramenta e pelo especialista são equivalentes e, portanto, a proposta de utilizar a sustentação do Pitch como métrica para reconhecimento da patologia mostrou-se eficiente...

Objectives: Develop an automated tool for recognition of segments without the presence of voice during phonation of the patient based on Pitch sustainment. Method: The procedures for construction and verification of the technique are the acquisition of voice, windowing, application of Discrete Fourier Transform, the Pitch detection and verification of Pitch. Results: With the analysis of 101 voices, the tool diagnosed 56 voices with laryngeal dystonia and 45 as healthy. Already the specialist diagnosed 53 voices with laryngeal dystonia and 48 voices as healthy. Conclusion: The results showed that the diagnosis made by the tool and specialist is equivalent and therefore the proposed use of the Pitch sustainment as a measure for the recognition of the pathology was effective...

Objetivos: Desarrollar una herramienta automatizada para el reconocimiento de segmentos sin la presencia de la voz durante la fonación basado en el apoyo de la Pitch. Método: Los procedimientos para la construcción y la verificación de la técnica son la adquisición de la voz, de ventanas, la aplicación de la Transformada Discreta de Fourier, la detección de tono y la verificación de Pitch. Resultados: Con el análisis de 101 voces, la herramienta ha diagnosticado 56 voces con distonía laríngea y 45 lo más saludable. Ya el especialista diagnostica 53 voces con distonía laríngea y 48 voces tan saludables. Conclusión: Los resultados mostraron que el diagnóstico realizado por la herramienta y por el especialista son equivalentes y por lo tanto el uso propuesto de la Pitch apoyo como una métrica para el reconocimiento de la patología fue eficiente...

Contribuições da Interface Usuário-Computador nos serviços hospitalares / Contributions of the User-Computer Interface in the hospital services / Contribuciones de la Interface Usuario-Ordenador en los servicios hospitalares

Objetivo: Este trabalho apresenta um breve levantamento sobre questões relacionadas à Interface Usuário-Computador, Usabilidade e Ergonomia, assim como suas aplicações em uma solução de Sistema de Informação na área hospitalar. Método: O experimento foi realizado com usuários de diferentes perfis demográficos e de familiaridade com computadores. Os participantes, funcionários do hospital, foram submetidos ao uso de duas interfaces gráficas, uma que seguia as características de um bom design de interface e a outra que apresentava deficiência em tal quesito. Resultados: Os resultados obtidos indicaram como contribuições de um bom design de interface, uma redução no tempo e na quantidade de erros na operação da máquina, que podem trazer ganhos como a redução dos custos e o aumento da qualidade dos serviços, melhorando a imagem da organização. Conclusão: Uma interface usuário-computador adequada implica diretamente no desempenho dos funcionários, organização e satisfação dos pacientes que são beneficiados de múltiplas formas em diversos processos.

Objective: This paper briefly develops theory about issues related to Human-Machine Interface, Usability and Ergonomics, as well as their applications to Hospital Information System. Methods: The experiment was conducted with different users concerning their demographic profile and familiarity with computers. Those participants, doctors, nurses and administrative staff, handled two graphical interfaces, one presenting good interface design and the other one without any orientation. Results: Results point to an important decrease in time and errors while operating the machine if a good interface design is implemented. This can bring reduced costs and more perceived service quality, improving the image of the organization. Conclusion: A proper User-Computer Interface implies directly in the employee performance, organization and patients satisfaction which are benefited in multiple ways in different processes.

Objetivos: Este trabajo presenta un breve levantamiento sobre cuestiones relacionadas a la Interface Usuario-Ordenador, la Usabilidad y la Ergonomía así como sus aplicaciones en una solución del Sistema de Información en el área del hospital. Métodos: El experimento fue llevado con los usuarios de diversos perfiles demográficos y la familiaridade con las computadoras. Los participantes, personal del hospital, fueron sometidos al uso de dos interfaces gráficas, una que siguió las características de un buen design del interface y la outra que presentaba deficiencia en tal quesito. Resultados: Los resultados indicaron como contribuciones de un buen design de interface una reducción en el tiempo y la cantidad de errores en la operación de la máquina, que puede aportar beneficios tales como la reducción de costos y aumento de la calidad de los servicios, mejorando la imagen de la organización. Conclusión: Una interface usuario-ordenador implica directamente el buen desempeño de los empleados, organización y la satisfacción de los pacientes que se benefician de muchas maneras en los diferentes procesos.