Article
Forestry
Andre R. de Geus, Andre R. Backes, Alexandre B. Gontijo, Giovanna H. Q. Albuquerque, Jefferson R. Souza
Summary: The wood industry is crucial in many countries, but illegal logging is a common issue for reducing costs and obtaining valuable wood species. Recognizing wood species is important to combat this, the study introduces a simplified acquisition process and deep learning models for wood classification, showing promising results.
WOOD SCIENCE AND TECHNOLOGY
(2021)
Article
Energy & Fuels
Thiago Averaldo Bimestre, Fellipe Sartori Silva, Celso Eduardo Tuna, Jose Carlos dos Santos, Joao Andrade de Carvalho Jr, Eliana Vieira Canettieri
Summary: This study investigates the physicochemical and thermogravimetric characteristics of 21 wood species from the Amazon region, as well as their higher heating value (HHV), focusing on the energy use of biomass. The samples showed a high lignin content and varying cellulose and hemicellulose contents, with different wood species having different combustion capabilities.
Article
Agriculture, Multidisciplinary
Anna Fabijanska, Malgorzata Danek, Joanna Barniak
Summary: This paper introduces a convolutional neural network method for automatic tree species identification from scanned wood core images, achieving high accuracy in wood patch classification and wood core classification tasks. The model outperformed the state-of-the-art methods and the study also analyzed the impact of model parameters and training settings on performance to ensure the highest recognition rates.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2021)
Article
Environmental Sciences
Ivo Augusto Lopes Magalhaes, Osmar Abilio de Carvalho Junior, Osmar Luiz Ferreira de Carvalho, Anesmar Olino de Albuquerque, Potira Meirelles Hermuche, Eder Renato Merino, Roberto Arnaldo Trancoso Gomes, Renato Fontes Guimaraes
Summary: This research mapped the vegetation in the state of Amapa within the Amazon biome using the phenological behavior of the Sentinel-1 time series. Different image sets and classifiers based on deep learning and machine learning were compared. The study showed that phenological variations based on long Synthetic Aperture Radar (SAR) time series allow for the accurate representation of land cover/land use and water dynamics.
Article
Ecology
Ismail Kirbas, Ahmet Cifci
Summary: This study investigates the classification of wood species using convolutional neural networks and evaluates the performance of various deep learning architectures. The experimental findings demonstrate that the Xception model achieves remarkable performance on the WOOD-AUTH dataset.
ECOLOGICAL INFORMATICS
(2022)
Article
Construction & Building Technology
Sohrab Rahimi, Vahid Nasir, Stavros Avramidis, Farrokh Sassani
Summary: Moisture monitoring is crucial during timber drying to maintain quality, and this study aims to estimate the final moisture distribution of kiln-dried timber using machine learning. Initial moisture level and weight were found to be essential variables affecting timber overdrying and underdrying. The decision tree approach showed better performance for moisture classification, with around 91% accuracy.
STRUCTURAL CONTROL & HEALTH MONITORING
(2022)
Article
Plant Sciences
Prabu Ravindran, Frank C. Owens, Adam C. Wade, Patricia Vega, Rolando Montenegro, Rubin Shmulsky, Alex C. Wiedenhoeft
Summary: Illegal logging poses a major threat to forests in Peru and globally. To address this issue, a convolutional neural network was trained using transfer learning for wood identification, achieving high accuracy and readiness for real-world field screening scenarios. This technology helps monitor, incentivize, and monetize legal and sustainable wood value chains.
FRONTIERS IN PLANT SCIENCE
(2021)
Article
Environmental Sciences
Marks Melo Moura, Luiz Eduardo Soares de Oliveira, Carlos Roberto Sanquetta, Alexis Bastos, Midhun Mohan, Ana Paula Dalla Corte
Summary: The study aimed to evaluate the use of high-resolution UAV images for identifying forest species in areas of forest regeneration in the Amazon. By training convolutional neural networks with different thresholds, the results showed that CNN can accurately identify species with over 90% accuracy. The study concluded that CNN is an effective tool for classifying species in UAV images.
Article
Green & Sustainable Science & Technology
Leticia da Silva Moreira, Fernando Wallase Carvalho Andrade, Bruno Monteiro Balboni, Victor Hugo Pereira Moutinho
Summary: This study compared the physical and mechanical characteristics of branch wood of three tree species with their respective stems. The study found no significant differences in basic density between branch and stem wood, but branch wood had lower coefficient of anisotropy compared to stem wood. D.odorata showed similar mechanical properties between branch and stem wood, while H. petraeum and H. courbaril branch wood had lower strengths compared to stem wood.
Article
Engineering, Biomedical
Zefang Lin, Weihong Yang, Wenqiang Zhang, Chao Jiang, Jing Chu, Jing Yang, Xiaoxu Yuan
Summary: This study aims to develop and evaluate a deep learning-based classification model for recognizing the pathology of renal tumors. The experiment showed satisfactory performance in distinguishing malignant and benign tumors, as well as recognizing subtypes of renal tumors.
BIOMEDICAL ENGINEERING ONLINE
(2023)
Article
Plant Sciences
Geovanni Figueroa-Mata, Erick Mata-Montero, Juan Carlos Valverde-Otarola, Dagoberto Arias-Aguilar, Nelson Zamora-Villalobos
Summary: Tree species identification is vital for conservation, sustainable management, and combating illegal logging. This research developed a CNN-based automated tree species identification system with high accuracy. Additionally, an Android application was developed to identify tree species from images of wood cross-sections.
FRONTIERS IN PLANT SCIENCE
(2022)
Article
Multidisciplinary Sciences
Aaron N. Shugar, B. Lee Drake, Greg Kelley
Summary: The innovative approach of combining X-ray fluorescence spectrometry with convolutional neural network machine learning allows for rapid and accurate identification of wood species, supporting environmental protection laws and regulations.
SCIENTIFIC REPORTS
(2021)
Article
Computer Science, Information Systems
Mao Xiao, Chun Guo, Guowei Shen, Yunhe Cui, Chaohui Jiang
Summary: This paper presents a malware classification method based on PE files, using a new visualization method and deep learning technology to improve the accuracy and efficiency of malware classification.
COMPUTERS & SECURITY
(2021)
Article
Forestry
Maryam Shirmohammadi
Summary: The antimicrobial characteristics of Australian commercial timber species were studied, and it was found that hardwood samples showed inhibition of microorganisms, while plastic and paper samples as well as softwood samples did not show significant differences. Further studies can explore the types of extractives in timber, their role in antimicrobial characteristics, and the differences in timber surface types related to microbial contamination.
Article
Forestry
Fanyou Wu, Rado Gazo, Eva Haviarova, Bedrich Benes
Summary: This research demonstrates the feasibility of using deep convolutional neural networks (CNNs) for hardwood lumber identification, achieving an accuracy rate of 98.2% in 11 common hardwood species classification tasks.
WOOD SCIENCE AND TECHNOLOGY
(2021)
Article
Forestry
Camila Costa de Seabra, Humberto Angelo, Alexandre Nascimento de Almeida, Joaquim Carlos Goncalez, Maristela Franchetti de Paula, Gislayne da Silva Goulart, Sandra Regina Afonso, Elisa Palhares de Souza, Alexandre Bahia Gontijo
Summary: The wood from the Caatinga biome in Brazil has unique color attributes that are considered important in adding value to products, making it a potential alternative to traditional woods in the market.
FOREST PRODUCTS JOURNAL
(2022)
Article
Materials Science, Multidisciplinary
Lays Furtado de Medeiros Souza Kataoka, Rosineide Miranda Leao, Alexandre Bahia Gontijo, Maria del Pilar Hidalgo Falla, Sandra Maria da Luz
Summary: This study focused on the sheet resistance and conductivity of nanocomposite films made from regenerated cellulose and silver nanoparticles, which showed potential for electronic devices. Chemical treatments were able to remove impurities from the fibers, increase crystallinity index, and improve thermal stability.
MATERIALS TODAY COMMUNICATIONS
(2022)
Article
Computer Science, Hardware & Architecture
Rodrigo Moreira, Larissa Ferreira Rodrigues Moreira, Flavio de Oliveira Silva
Summary: The Internet is crucial for global applications and businesses, but security is a major challenge. The Darknet, a parallel network within the Internet, requires real-time classification due to malicious activities. Our paper proposes a novel approach using CNN and RL techniques for intelligent and adaptive packet sampling rates in high-performance networks. With a TOR traffic prediction accuracy of 99.84%, our method shows successful classification in high-throughput networks.
COMPUTERS & ELECTRICAL ENGINEERING
(2023)
Article
Computer Science, Information Systems
Diego Gabriel Soares Pivoto, Tiberio Tavares Rezende, Michelle Soares Pereira Facina, Rodrigo Moreira, Flavio de Oliveira Silva, Kleber Vieira Cardoso, Sand Luz Correa, Antonia Vanessa Dias Araujo, Rogerio Sousa E. Silva, Heitor Scalco Neto, Gustavo Rodrigues de Lima Tejerina, Antonio Marcos Alberti
Summary: As society evolves, new demands arise, requiring significant effort to be met. This article aims to create a methodology for analyzing the relevance of enabling technologies and use it to design an optimal 6G architecture capable of meeting the demands. Two methods, AVG and AHP, are selected to determine the relevance of enablers and classify an optimal set of enabling technologies for a 6G architecture.
Article
Forestry
Laise de Jesus dos Santos, Lohana Vieira Souza, Gabriele Melo de Andrade, Thayrine Silva Matos, Marcelo Mendes Braga Junior, Mirtes Emilia Almeida Manacas, Alexandre Bahia Gontijo, Joao Carlos Ferreira D. E. M. E. L. O. Junior, Javan Pereira Motta, Luiz Eduardo de Lima Melo
Summary: This study investigates the wood usage by two indigenous ethnic groups in the Brazilian Amazon and identifies the corresponding species. It finds that these groups prefer medium density wood for construction and high-density wood for crafting hunting and warfare artifacts. However, the increasing environmental threats pose a challenge to the preservation of the cultural heritage of indigenous peoples.
Proceedings Paper
Computer Science, Information Systems
Rodrigo Moreira, Hugo G. V. O. da Cunha, Larissa F. Rodrigues Moreira, Flavio de Oliveira Silva
Summary: This study proposes an intelligent method called VINEVI for seamless monitoring of heterogeneous infrastructures and applications. The VINEVI architecture combines real-time traffic classification with well-known tools to monitor the entire stack from hardware to virtualized applications, achieving advanced level of detail monitoring for heterogeneous infrastructures.
ADVANCED INFORMATION NETWORKING AND APPLICATIONS, AINA-2022, VOL 1
(2022)
Proceedings Paper
Computer Science, Information Systems
Eduardo Castilho Rosa, Flavio de Oliveira Silva
Summary: This article reviews and compares recent implementations of FIB in physical switches, providing new insights and future research directions in this field.
ADVANCED INFORMATION NETWORKING AND APPLICATIONS, AINA-2022, VOL 3
(2022)
Proceedings Paper
Computer Science, Information Systems
Diego Nunes Molinos, Romerson Deiny Oliveira, Marcelo Silva Freitas, Natal Vieira de Souza Neto, Marcelo Barros de Almeida, Flavio de Oliveira Silva, Pedro Frosi Rosa
Summary: The Internet's evolution is limited, requiring improved QoS and QoE. SDN is used to redesign the architecture and address TCP/IP issues. This study presents a specification and development of an application-driven SDN switch, allowing reconfiguration of network behavior through the control plane.
ADVANCED INFORMATION NETWORKING AND APPLICATIONS, AINA-2022, VOL 2
(2022)
Article
Computer Science, Hardware & Architecture
Thiago Bueno da Silva, Fabio L. Verdi, Juliano Coelho Goncalves de Melo, Jose Augusto Suruagy, Flavio Silva, Antonio M. Alberti
Summary: The Internet has transformed our interactions with the world, but it also comes with limitations. To address these challenges, researchers have proposed future Internet architectures (FIAs), such as NovaGenesis (NG). This article presents the NG data and control planes, and their application in a P4-based future Internet exchange point.
Proceedings Paper
Computer Science, Interdisciplinary Applications
Rodrigo Elias Francisco, Flavin de Oliveira Silva
Summary: The demand for qualified professionals in software maintenance poses challenges to computer education. This study presents an Expert Knowledge Module (EKM) and a content recommendation engine using AI techniques to provide adequate support in teaching. Simulation experiments validate the effectiveness of the recommendation mechanism.
CSEDU: PROCEEDINGS OF THE 14TH INTERNATIONAL CONFERENCE ON COMPUTER SUPPORTED EDUCATION - VOL 1
(2022)
Proceedings Paper
Computer Science, Interdisciplinary Applications
Rodrigo Elias Francisco, Flavin de Oliveira Silva
Summary: Intelligent Tutoring Systems (ITS) in computer science education are automated and adaptive technologies that address the challenges students face in dealing with problem complexity and learning existing technologies. This study investigates the development and application of ITS in computer science education, with a focus on assessing AI techniques, algorithms, and datasets. The results highlight challenges in research, such as dataset unavailability and reproducibility difficulties, the lack of in-depth explanations of the relationship between AI techniques and ITS data, the need for further exploration of AI techniques, and the importance of research in software engineering for ITS.
CSEDU: PROCEEDINGS OF THE 14TH INTERNATIONAL CONFERENCE ON COMPUTER SUPPORTED EDUCATION - VOL 1
(2022)
Proceedings Paper
Computer Science, Information Systems
Eduardo Castilho Rosa, Flavin de Oliveira Silva
Summary: One core component of Named-Data Networking (NDN) is FIB, which uses Longest Name Prefix Matching (LNPM) to route hierarchical names. This work proposes a hash-free FIB model for programmable data planes, which balances lookup speed and memory consumption by segregating FIB and using packet recirculations for LNPM.
36TH INTERNATIONAL CONFERENCE ON INFORMATION NETWORKING (ICOIN 2022)
(2022)
Proceedings Paper
Computer Science, Theory & Methods
Vandre Leal Candido, Flavio de Oliveira Silva
Summary: IoT requires management of devices with interoperability and reliability as key issues, lacking a tool with these characteristics in current literature. This work introduces a tool that achieves interactive job execution over a network of heterogeneous devices with reliable message delivery, leveraging the power of AMQP protocol and RabbitMQ message broker. The tool's architecture and evaluation show its ability to offer interactivity and reliable message delivery under different scenarios.
CLOSER: PROCEEDINGS OF THE 11TH INTERNATIONAL CONFERENCE ON CLOUD COMPUTING AND SERVICES SCIENCE
(2021)
Proceedings Paper
Computer Science, Theory & Methods
Daniel Ricardo Cunha Oliveira, Mauricio Amaral Goncalves, Natal Vieira de Souza Neto, Flavio de Oliveira Silva, Pedro Frosi Rosa
Summary: The article discusses the modeling, implementation, and performance of a network self-configuration system for meeting the specialized configuration demands of new edge elements. The system, based on the SONAr framework, is capable of implementing specific network segment parameterization and acting as a specialized plug-and-play feature.
CLOSER: PROCEEDINGS OF THE 11TH INTERNATIONAL CONFERENCE ON CLOUD COMPUTING AND SERVICES SCIENCE
(2021)
Proceedings Paper
Computer Science, Information Systems
Rodrigo Moreira, Larissa Ferreira Rodrigues, Pedro Frosi Rosa, Rui L. Aguiar, Flavio de Oliveira Silva
Summary: This paper introduces the development of network slicing technology to meet different user requirements and efficiently provide customized resources, as well as a method of guiding path-setting agents through convolutional neural networks. Experimental results demonstrate the suitability of convolutional neural networks for enhancing network slicing and guiding NASOR to establish network slices over multiple domains.
35TH INTERNATIONAL CONFERENCE ON INFORMATION NETWORKING (ICOIN 2021)
(2021)