Brazil

Inspection of weld radiographs is an essential task for preventing leaks in oil industry pipelines. Automatic inspection is important to assist specialists whose performance can be influenced by fatigue, visual acuity, and experience. A challenging issue in automatic inspection is the availability of laudable images that can be used for training classifiers. Data augmentation is an alternative however it can produce images with little or even negative impact on the training. This article presents a method for filtering images, aiming to favor the classifier relearning process, which is done by eliminating images with low performance from the training set and further retraining the segmentation model, with a positive impact on its generalization capabilities. Experimental results of the method show that data filtering accelerates the learning process and produces an improvement in the segmentation performance.


BRACIS 2020

Training Data Filtering for Deep Learning Applied toInspection of Welded Joints in Oil Pipelines

artificial neural networks

computer vision

deep learning

AIAI 2020 (16th International Conference on Artificial Intelligence Applications and Innovations) should be held in Halkidiki, Greece. Due to current Covid-19 situation the conference will be held virtually. Underline will host a repository of video talks and preserve them for the future. Videos will be released on Underline one day after the conference.

AIAI 2020

technical paper

The 9th Brazilian Conference on Intelligent Systems (BRACIS), the 16h Encontro Nacional de Inteligência Artificial e Computacional (ENIAC), and the 12th Concurso de Teses e Dissertações em Inteligência Artificial e Computacional (CTDIAC) was held from 20 to 23 October, 2020.

BRACIS originated from the combination of the two most important scientific events in Brazil in Artificial Intelligence (AI) and Computational Intelligence (CI): the Brazilian Symposium on Artificial Intelligence - SBIA (22 editions), and the Brazilian Symposium on Neural Networks - SBRN (13 editions). ENIAC is a national forum for researchers, practitioners, educators and students to present and discuss innovations, trends, experiences, developments and work in progress in the fields of Artificial and Computational Intelligence. CTDIAC is a satellite event whose goals are to disseminate and award the best dissertations and theses (MSc and PhDs concluded in Brazilian universities from June 2018 to May 2020) in Artificial and Computational Intelligence areas.

In this year, the Symposium on Knowledge Discovery, Mining and Learning (KDMiLe) will be held in conjunction with BRACIS 2020. This sysposium aims at integrating researchers, practitioners, developers, students and users to present their research results, to discuss ideas, and to exchange techniques, tools, and practical experiences – related to the Data Mining and Machine Learning areas.

A previsão de receita tributária é uma tarefa desafiadora às secretarias de fazenda. Por conta disso, neste trabalho se inicia um projeto de investigação do uso do modelo LSTM para a previsão de receita tributária, em especial o tributo ICMS. As previsões obtidas pelo modelo LSTM são comparadas às realizadas pela SEFAZ e pelo modelo de rede neural MLP. A rede LSTM apresentou resultados promissores para a previsão do tributo ICMS.

Previsão de Arrecadação do Tributo ICMS do Estado do Rio de Janeiro Utilizando a Rede Neural LSTM

Deep Learning consists of modern techniques that approach one or more methods of artificial intelligence. An approach is to use Convolutional Neural Networks together with traditional Neural Networks for digital image processing. In this work, a research is conducted to evaluate a deep learning technique in the classification of potassium (k) macronutrient deficiency by the soybean leaf. This research presents an own dataset with distinct treatments of the potassium macronutrient. Several deep learning scenarios are evaluated with different metrics. The results are compared with the literature and show a great potential of Convolutional Neural Networks, with precision above 99% in this type of classification.

Classifying the Macronutrient Deficiency in Soybean Leaf with Deep Learning

In this work was proposed an approach to classify the organizational structure of a group of moving target agents that are continuously monitored by a smaller group of moving observer agents. The approach considers that the group of target agents can organize themselves according to eight different paradigms. The agents communicate through the exchange of messages whose contents are speech-act performative. We evaluate the approach considering seven techniques to solve the problem of classifying the group of target agents' organizational structure.

Classifying Organizational Structures on Targets in the Cooperative Target Observation

Currently, keyboards are the most common means of communicating with computers. Despite being the most commonly used keyboard layout, QWERTY has had various issues raised concerning its effectiveness, as it is not efficient in English (target language) or in fact other languages. Therefore, this paper presents the development process of a Genetic Algorithm with the intention of generating a more adequate and coherent layout proposal for Brazilian Portuguese, which has its focus on ergonomics and user productivity. Using five ergonomic criteria and a statistical analysis of the characters and sequences of most frequently used pairs in Brazilian Portuguese, a layout approximately 53% better than QWERTY was obtained.

Using Genetic Algorithms to Design an Optimized Keyboard Layout for Brazilian Portuguese

Convolutional Neural Networks (CNNs) is the main solution for image classification tasks in different applications. However, selecting the most suitable CNN and its parameters for a given image dataset is usually performed by trial and error, which may take much time and computational cost. This paper proposes the Dataset2Vec method and employs Meta-Learning (MtL) to select CNN architectures for image classification. Dataset2Vec adopts a deep neural network to extract features from images datasets, embedding them in a single feature vector. To evaluate the proposed solution, it was adopted to select among six CNN algorithms for 45 two-classes image datasets. The results showed that the MtL using Dataset2Vec outperformed by different baseline methods in all performance measures evaluated, indicating the proposal was able to extract representative features from datasets of images for CNN selection.

Meta-Characteristics Extraction from Image Datasets for Selection of Convolutional Neural Networks

The classification and separation of documents is a crucial step in the
analysis of court lawsuits. Deep learning algorithms have achieved promising
results in this task, extracting relevant information from the texts of these documents. However, court documents have become increasingly heterogeneous,
i.e., photos, receipts, text documents, etc., directly impacting classification accuracy. This work investigates the use of multimodal convolutional networks,
combining characteristics extracted from texts and images to classify court lawsuits pages. Two multimodal approaches were compared with four single-mode.
In terms of accuracy and kappa, all algorithms were evaluated in a database
composed of 117 lawsuits. The results showed that the approach that achieved
the best performance is multimodal, presenting effectiveness and efficiency in
the classification of court lawsuits pages.


Classification of Court Lawsuits Pages using MultimodalConvolution Neural Networks

Identifying subjective sentences and calculating the polarity of these sentences are two important tasks of sentiment analysis. Although they are not new problems, the most of solutions are focused on the English language. In this paper, we have proposed and evaluated an approach based on machine learning to handle both tasks in Portuguese. We investigated the use of two classifiers and also proposed a set of linguistic features extracted from the text. We experiment and evaluate the methods against a representative set of baselines and a diverse set of datasets. Our approach achieved the best results for both tasks and for all test datasets.

Investigating Sentences Features for Subjectivity and Polarity Classification in Brazilian Portuguese

Displasia Cortical Focal (DCF) é um tipo de lesão cerebral que é a principal causa de Epilepsia Refratária em crianças, e o terceira maior causa em adultos, estando presente em mais de 50% dos casos infantis e 20% nos casos adultos. Pacientes com essa doença sofrem de crises epiléticas onde os remédios não são capazes de ajudar. O tratamento clínico mais indicado nesse caso é a cirurgia. Mas para ser realizada é necessário se ter uma identificação precisa da lesão. Esse processo é bastante complexo pois a região da lesão não é bem definida para visualização em Imagens de Ressonância Magnética (IRM). Porém com os avanços computacionais sugiram várias técnicas para o auxílio na classificação de imagens médicas. Neste trabalho é estudado o uso de Redes Neurais Convolucionais (CNN) para classificação de imagens com a presença da DCF. Foram utilizadas imagens no plano axial de 7 pacientes previamente diagnosticados com DCF. De cada paciente foram obtidos vários cortes e posteriormente cada corte foi divido em janelas de classificação. Estas janelas por sua vez foram utilizadas para realizar o treinamento e análise dos resultados obtidos da CNN. Os resultados obtidos nesse trabalho se mostraram bastante satisfatórios para classificação de janelas com DCF, porém a rede apresentou algumas dificuldades para classificar as janelas sem DCF. Com isso seu uso para demarcação em uma imagem completa ainda não se mostrou adequado. As limitações e novas abordagens são discutidas.

Use of Convolutional Neural Networks to Identify Focal Cortical Dysplasia in Patients with Refractory Epilepsy

Sports analytics, also known as scout, has gained significant atten-tion thanks to the positive results of its application in a wide variety of sports.Due to the popularity of football, several studies seek to apply Machine Lear-ning on scout data in this sport due to the popularity of soccer. This type ofdata comprises events countings, such as passes and finishes, and can assist thetechnical staffs’ decision-making process. However, these efforts are focused onEuropean football. In this work, we investigate the knowledge discovery usingMachine Learning algorithms in data from scout obtained in Brazilian soccer.With that, we show the current potential and limitations of this approach.

Knowledge Discovery in Brazilian Soccer Championship Scout Data

To guarantee a high indoor air quality is an increasingly important task. Sensors measure pollutants in the air and allow for monitoring and controlling air quality. However, all sensors are susceptible to failures, either permanent or transitory, that can yield incorrect readings. Automatically detecting such faulty readings is therefore crucial to guarantee sensors' reliability. In this paper we evaluate three Machine Learning algorithms applied to the task of classifying a single reading from a sensor as faulty or not, comparing them to standard statistical approaches. We show that all tested machine learning methods -- Multi-layer Perceptron, K-Nearest Neighbor and Random Forest -- outperform their statistical counterparts, both by allowing better separation boundaries and by allowing for the use of contextual information. We further show that this result does not depend on the amount of data, but ML methods are able to continue to improve as more data is made available.

Training Data Filtering for Deep Learning Applied toInspection of Welded Joints in Oil Pipelines

Downloads

Next from BRACIS 2020

Previsão de Arrecadação do Tributo ICMS do Estado do Rio de Janeiro Utilizando a Rede Neural LSTM

Similar lecture

Introducing an edge-native deep learning platform for exergames

Training Data Filtering for Deep Learning Applied toInspection of Welded Joints in Oil Pipelines

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from BRACIS 2020

Previsão de Arrecadação do Tributo ICMS do Estado do Rio de Janeiro Utilizando a Rede Neural LSTM

Similar lecture

Introducing an edge-native deep learning platform for exergames

Downloads