Machine learning tools are widely used in support of bioacoustics studies, and there numerous publications on the applicability convolutional neural networks (CNNs) to automated presence-absence detection species. However, relation between merit acoustic background modelling recognition performance needs be better understood. In this study, we investigated influence substance detector White-lored Spinetail (Synallaxis albilora). Two designs were evaluated: 152-layer ResNet with transfer a...

Purpose As in the private sector, public organizational information systems (IS) development is commonly carried out through projects. One of alternatives followed by governmental organizations to perform their projects outsourcing (by hiring other institutions that have expertise IS area be developed). However, limited research has been conducted on project success regarding these government-to-government (G2G) contexts. Since achieving crucial for management, this paper proposes a model...

The public sector has many challenges when improving business processes. One of those is managing the projects' success by considering context and constraints management process (BPM). Moreover, importance elements that will drive can be different depending on stakeholder stage project (e.g. initiation, executing, closing), which adds another ingredient complexity to management. Aiming give capacity manage criteria factors in BPM projects, this research used action develop an integrated...

Meteorological elements can affect the environment and cultures differently may alter natural development process contributing significantly to climate change. variables of Brazilian Pantanal were studied used determine evapotranspiration with fewer variables. It was found that artificial intelligence substantially improve environmental modeling when alternative prediction techniques are used, resulting in lower project costs more reliable results. This work tried find best combination by...

Automated acoustic recognition of birds is considered an important technology in support biodiversity monitoring and conservation activities. These activities require processing large amounts soundscape recordings. Typically, recordings are transformed to a number features, machine learning method used build models recognize the sound events interest. The main problem scalability data processing, either for developing or made over long time periods. In those cases, resources required might...

Rainfall is the key element in regional water balance, and has direct influence over economic activity. Quantifying rainfall at spatial temporal scales regions where meteorological stations are scarce important for agriculture, natural resource management land-atmosphere interactions science. Thus, we evaluated neural network performance estimates Mato Grosso State located Brazilian Midwest region. A dataset of 12 was used to train network, then run perform estimates, which allowed comparing...

Os dados meteorológicos são de grande importância para os estudos científicos, pois auxiliam na tomada decisões em diferentes áreas do conhecimento. As estações automáticas realizam o trabalho captar esses dados, porém, problemas podem ocorrer nos instrumentos causando falhas nas séries e inutilizando um período ou até mesmo toda a série. Visto que análise desses é prejudicada com esse problema, as devem ser tratadas garantir uma maior qualidade obtenção das informações. Este visa comparar...

Researches involving the long-wave balance have intensified in order to understand environmental performance of micrometeorological variables. This study tried determine parameters Brunt’s equation for northern Pantanal Mato Grosso, using “genetic algorithm”. The results show that values estimated by model significantly approached obtained with data collected locally. It was also noted there a seasonal variation these parameters, and Brunt adjusted rainy dry season use parameter (a) (b)...

This paper presents an approach for the detection of highway guardrails using camera-based systems and advanced machine learning techniques. The proposed methodology combines feature extraction with Convolutional Neural Networks (CNN), specifically MobileNetV2, ResNet18, VGG16, clustering algorithms applied to these features. effectiveness models is evaluated through classification metrics, a particular emphasis on Gaussian Mixture Model (GMM) forming more cohesive well-separated clusters...

Geolocation methods identify objects in images and determine their geospatial locations. Current object geolocalization face challenges such as high hardware costs, limited class coverage, difficulties with repeated occurrences, performance issues dynamic environments. This paper introduces a machine learning approach for geolocalizing from low frame rate video using single camera image metadata, aiming to reduce costs complexity compared traditional methods. The method combines displacement...

Esse trabalho aborda a identificação automática de algumas espécies orquídeas da região Chapada dos Guimarães, utilizando técnicas classificação imagens com Inteligência Artificial, especificamente Deep Learning. Para isso, foi criada uma base dados coletadas internet. A técnica data augmentation aplicada para enriquecer o conjunto treinamento e evitar overfitting. Dentre os modelos testados, rede neural convolucional arquitetura 32-6416-64 atingiu acurácia cerca 65% teste, um resultado...

Artificial Neural Networks (ANN) have been widely used to model several types of data. The precision ANN models is dependent upon their configuration, i.e., input parameters, training algorithm and architecture configurations. problem lies in the amount possible combinations these parameters which results countless unique ANNs. One method finding a good combination use Genetic Algorithm (GA). Several studies combine GA with an solve problems, however, it not clear should determine. This work...

Abstract This paper presents Mannga (Multiple variables with Artificial Neural Network and Genetic Algorithm), a method designed for gap filling meteorological data. The main approach is to estimate the missing data based on values of other measured at same time in local, since are strongly related. Experimental tests showed performance compared two methods typically used by researches this area. Good results were achieved, high accuracy even sequential failures, which big challenge...

As estações tradicionais de coleta dados meteorológicos estão sendo substituídas por automatizadas, gerando uma maior quantidade aquisição informações. Entretanto, há também a probabilidade erros como ausência dados, o que gera dificuldades na análise das pesquisas e tomada decisão dos processos. Tais falhas nos devem ser corrigidas para se obter base qualidade possibilitar análises mais detalhadas conclusivas. Existem métodos preenchimento realizam este trabalho. No entanto, cada método tem...

Diversos trabalhos foram realizados para o reconhecimento de COVID-19 por meio imagens raio-X. Os obtinham bom desempenho no das imagens, entanto, os modelos estão alinhados aos conjuntos dados utilizados, que não implica mesmo fora do contexto treino. Deste modo, este trabalho aplicou uma forma justa avaliar em diferentes cenários. resultados demonstraram conseguiram distinguir entre origem diferente, assim, foi determinado estiveram adaptados ao conjunto obtido.