Luis Enrique Raya-González, Víctor Alfonso Alcántar-Camarena, Jonathan Cepeda-Negrete
et al.
Manual monitoring of pests and diseases in maize crops requires considerable time and resources, significantly increasing production costs. Artificial intelligence (AI)-based studies have explored their automated detection, primarily through transfer learning architectures, although with limited success. This study evaluated and compared four AI approaches: convolutional neural networks (CNN), a hybrid CNN with support vector machines (CNN-SVM), mixture of experts (MoE) models, and transfer learning architectures. Eighteen CNN models were developed and optimized using a factorial design, and the best-performing model was used as the foundation for constructing the hybrid CNN-SVM and CNN-SVM-MoE models. The CNN-SVM-MoE model achieved the highest accuracy (99.14 %) and demonstrated strong generalization capabilities, even with data collected under field conditions. In contrast, transfer learning architectures showed lower performance. Statistical analysis revealed significant differences among the models, highlighting the superiority of the CNN-SVM-MoE approach. The results confirm that MoE models enhance performance in classifying maize pests and diseases and offer strong potential for integration into mobile or embedded devices, enabling their direct application in the field.
Mampi Devi, Sarat Saharia, Dhruba Kumar Bhattacharyya
et al.
Abstract To digitize and preserve the cultural heritage in the form of Indian classical dance become apparent area of research. Sattriya classical dance of North-East India (Assam) is one of the eight Indian classical dance forms that requires immediate preservation. Sattriya classical dance consists of 29 Asamyukta hastas (single-hand gestures) and 14 Samyukta hastas (double-hand gestures). Moreover, the foundation of Samyukta hasta depends on understanding Asamyukta hasta. Therefore, the paper aims to classify single-hand gestures of Sattriya classical dance only. Although, a solution based on two level classification method to classify the Sattriya classical dance is available in recent literature, but it requires a trial and error method to select the optimized features. Since, Asamyukta hastas can appear closely similar to each other and therefore misclassification chances are very high. Thus, accuracy rate obtained for the two level classification method was only 75.45%. So, to address this issues in this paper, a Multilevel Classification Model with Vision based Features (MCM- $$V_b$$ V b F) has been proposed to classify the Asamyukta hastas of Sattriya classical dance. This model uses two types of feature matching, viz., high-level feature matching and low-level feature matching. To extract the high-level features and low-level features different algorithm has been proposed. In this model, features are automatically selected. This proposed MCM- $$V_b$$ V b F model is also tested on Asamyukta hasta mudras of Bharatanatyam classical dance of South India (Tamil Nadu). This model obtain an accuracy 94.12%, 87.14% for Sattriya classical dance Single-Hand Gestures (SSHG) dataset and Bharatnatyam classical dance Single-Hand Gestures (BHSG) dataset respectively. This paper also provides the comparative study of the proposed model MCM- $$V_b$$ V b F with traditional bench-mark classifier model such as Naive Bayes, Decision Tree and Support Vector Classifier (SVM) etc.
Electric motorcycles are widely used due to their economic, portable, and easy-to-use characteristics. Power batteries are the primary power source of electric motorcycles. Electric motorcycles are usually pushed into elevators and parked at home or in enclosed corridor spaces for charging, which may pose serious safety hazards due to using inferior or expired batteries. The traditional manual management method is limited by human resources, making it difficult to manage and monitor such behavior. Automated detection of electric motorcycles based on artificial intelligence technology is an effective solution. Considering that common monitoring systems typically have limited data processing capabilities, this study proposes an electric motorcycle detection model based on improved You Only Look Once version 5s (YOLOv5s). Firstly, we develop the model by adding a transformer encoder module to the backbone of classical YOLOv5s. Next, the Bidirectional Feature Pyramid Network (BiFPN) is used for cross-scale connectivity and multiscale feature fusion. Finally, the Coordinate Attention module (CA) is added to improve the representation capacity of the target features and enhance the detection accuracy. The results of comparative experiments and ablation experiments verified the effective performance of the proposed model, which attained a mean average precision of 81.2%. Compared to classical models like faster R-CNN and YOLOv5, this methodology achieves higher performance with fewer parameters and computational complexity, meeting real-time requirements.
The Nex-Hys Project, funded by the German Federal Ministry of Economy as part of the WIPANO frame program, aimed to develop standards for safety characteristics of hybrid mixtures. These should be designed in a way that the standards are extended and the existing values in the experimental setup still hold true.
As part of the project, the Otto-von-Guericke University focused on the Minimum Ignition Temperature. Starting from a comparison of apparatuses for liquids (EN 14522) and dusts (IEC 80079-20-2) the decision to use the Godbert Greenwald Oven was made, as only there the possibility to introduce substances in gaseous and solid phase is given in principle.
The standardized setup was extended by a heated chamber to evaporate the liquids to be tested. Burnable gases were introduced into the system by means of mixing according to the partial pressure method in the pressurized air chamber or the evaporation chamber. Thus, it was possible to determine the MIT of the pure gases and vapors in the GG oven. Series of test for different substances under variation of the concentration show comparable values to the standard methods and to values found in literature. A trend to slightly higher temperatures (for gases and vapors) with a deviation close to the measurement uncertainty can be found. All in all, the deviation by different operators seems to be in the same range and cannot be neglected.
Following various combinations of dust, vapors and gases were tested. Up to now, no combinatory effect was detected. This seems to be in contradiction to former own publications, but there the setup was different. This could be an explanation for these nonreproducible result.
In all combinations, the MIT of the substances that has the lower MIT is dominating the final value. The setup-up generally proved to be able to test for the MIT of hybrid mixtures without violating existing values.
Part of the tests were made in a GG oven of double length showing a clear influence on the MIT for some substances, due to the longer residence time.
Apart from some rather practical weaknesses of the standardized setup there are general disadvantages that limit its use for a further development. First there is the totally unknow concentration and distribution of fuels in this open setup. Secondly, the subjective detection of the ignition by the operator. Third, the temperature distribution and heat transfer conditions that are not well defined.
The original intention behind the experiment is not in line with the scientific intention to create a versatile instrument to determine the ignition temperature for all phases and their mixtures. Therefore, a completely new approach would be necessary. The existing MIT standards for dust and vapor/gases can further exist unchanged and fulfil their purpose.
Chemical engineering, Computer engineering. Computer hardware
Abstract To perform a quantum brute force attack on a cryptosystem based on Grover's algorithm, it is necessary to implement a quantum circuit of the cryptographic algorithm. Therefore, an efficient quantum circuit design of a given cryptographic algorithm is essential, especially in terms of quantum security analysis, and it is well known that T‐depth should be reduced for time complexity efficiency. In this paper, the authors propose a novel technique to reduce T‐depth (and T‐count) when some quantum circuits located in between two Toffoli‐gates are interchangeable with a controlled phase gate (CP gate), and the authors apply this technique to five types of quantum adders, reducing T‐depth by more than 33%. The authors also present new SHA‐256 quantum circuits which have a critical path with only three quantum adders while the critical paths of quantum circuits in the previous studies consist of seven or 10 quantum adders, and the authors also apply our technique to the proposed SHA‐256 quantum circuits. Four versions of SHA‐256 quantum circuit are presented. Among the previous results, T‐depth of the circuit with the smallest Width (the number of qubits) 801 was approximately 109,104. On the other hand, T‐depth of the proposed SHA‐256 quantum circuit with the Width 797 is 16,055, which is remarkably reduced by about 85%. Another proposed quantum circuit only requires 768 qubits, which is the smallest Width compared to the previous results to the best of our knowledge. Furthermore, one other version is the most time‐efficient circuit with an overall Toffoli‐depth (and T‐depth) that is less than 5000.
Agriculture has a good stake in the world’s GDP. In many countries, agriculture and allied sectors have a good stake in national GDP. This paper covers details related to livestock since 1960s. The workforce has managed livestock for many decades. The workforce increases as the number of animals increases; it is an energy, time-consuming, and economically costly approach. Apart from it, there is no assurance about animal welfare in case of diseases, breeding, and feed intake issues. In the 21st century of digitalization, technology has a key role in improving overall monitoring, controlling, and processing in livestock management. This paper has gone thoroughly into the manual and automated livestock farm management, aiming welfare of animals, livestock products, consumers’ benefit, and sustainable environmental approaches.
I Gede Bawa Aprilyanta, Ariesta Lestari, Sherly Christina
Pharmacy is a business-oriented business which directly sells medicines to consumers. In ensuring that the drug supply is in accordance with market demand, the leadership of the Hasanah Palangka Raya Pharmacy usually estimates the drug supply for the future period based on sales reports and experience. For research purposes, it will provide a solution by developing Forecasting Applications by applying the Weighted Moving Average method with weights 2, 3, 4, 5, 6, 7, 8 and 9 periods and the Single Exponential Smoothing method with constant smoothing or (alpha), namely 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8 and 0.9. The results of the study resulted in the Weighted Moving Average method as the method that has the smallest Error value where from the 4 drugs that were used for forecasting, 3 drugs showed the Weighted Moving Average method as the best method which has the smallest Mean Absolute Deviation value and the smallest Mean Absolute Percentage Error value compared to Single Exponential Smoothing.
WANG Jinsong, YANG Weizheng, ZHAO Zening, WEI Jiajia
Blockchain technology has been widely used in finance, public services, the Internet of Things(IoT), network security, supply chains, and other fields.However, the traditional blockchain with a single chain structure has some deficiencies in throughput, transaction confirmation speed, and scalability, which makes it difficult to apply it in some short-term and high concurrency data scenarios.In this paper, the Directed Acyclic Graph(DAG) based blockchain technology has attracted extensive attention and studied by scholars because of its advantages, such as concurrent transaction confirmation, high throughput, and strong scalability.By analyzing and studying the development and evolution, evaluation methods, optimization direction, and application scenarios of the existing DAG based blockchains, this paper explores the feasibility of DAG based blockchains in landing applications.Through the development of a mainstream DAG based blockchain, it compares the advantages and disadvantages of traditional blockchains and DAG based blockchains, analyzes the existing blockchain attribute evaluation methods, and summarizes the current DAG based blockchain evaluation results.On this basis, this paper summarizes the optimization methods of the existing DAG based blockchain from the aspects of transaction confirmation speed, system throughput, system security, and storage structure, and summarizes the application of a DAG based blockchain in data management, data sharing based on edge computing and federated learning, and data security for access control and privacy protection.Finally, it points out the main problems and challenges in the current studies, and provides further research directions.
Mirza Hussein Sabki, Pei Ying Ong, Chew Tin Lee
et al.
Sustainable agriculture is an ongoing research strives for meeting society’s current food demand without compromising the future need and development. Maintaining soil fertility for quality farming is one of the essential parts. However, the wide application of synthetic agrochemicals (e.g., chemical fertiliser) has been a significant contributor to environmental pollution. This review aims to assess the potential of Rhodopseudomonas palustris (R. palustris), a purple non-sulphur bacterium, as a commercialised bio-fertiliser to sustainably promote plant growth. R. palustris is evaluated based on two defined pillars of sustainability, including the effects on plant growth, environmental impact, and feasible production. The effectiveness is based on the improvement of plant growth through the secretion of extracellular metabolites, resistance to abiotic stresses, bioremediation of heavy metals, and mitigation of greenhouse gas emissions. This review suggests the imperative roles of R. palustris as an effective bio-fertiliser in agriculture. However, the scalability of production and application deserved more attention. The potential substrates ranging from different waste streams and formulation methods for R. palustris production are summarised to discuss environmental and economic sustainability.
Chemical engineering, Computer engineering. Computer hardware
Abstract Data mining techniques were used to investigate the use of knowledge extraction in predicting customer churn in insurance companies. Data were included from a health insurance company for providing insight into churn behaviour based on a design and application of a prediction model. Additionally, three promising data mining techniques were identified for the prediction of modeling, including logistic regression, neural network, and K-means. The decision tree method was used in the modeling phase of CRISP-DM for identifying the attributes of churned customers. The predictive analysis task is undertaken through classification and regression techniques. K-means clustering variation is selected for exploring if the clustering algorithms categorize the customers in churning and non-churning groups with homogeneous profiles. The findings of the study show that data mining procedures can be very successful in extracting hidden information and get to know customer's information. The 50:50 training set distribution resulted in effective outcomes when the logistic regression technique was used throughout this study. A 70:30 distribution worked effectively for the neural network technique. In this regard, it is concluded that each technique works effectively with a different training set distribution. The predicted findings can have direct implications for the marketing department of the selected insurance company, whereas the models are anticipated to be readily applicable in other environments via this data mining approach. This study has shown that the prediction models can be utilized throughout a health insurance company's marketing strategy and in a general academic context with a combination of a research-based emphasis with a business problem-solving approach.
Computer engineering. Computer hardware, Information technology
Victor Antonio Figueroa Castillo, Carlos Andrés Villacreses Parrales, Jennifer Elizabeth Chóez Calle
et al.
La presente investigación tiene como finalidad identificar el potencial que tiene la tecnología blockchain y los contratos inteligente, todo esto debido a la creciente demanda de mayor transparencia en la administración pública a nivel mundial, en el cual se propone que los datos sean públicos, además del establecimiento de otros mecanismos, así como seguir utilizando los aportes de nuevos procesos para incrementar la capacidad de gestión, enfocándose en mejores controles y mecanismos de gobernanza. Una de las nuevas tecnologías que presenta potencial para ser utilizada en la protección de las organizaciones ante la corrupción es el blockchain, que en los actuales momentos varias compañías y gobiernos la están utilizando. Un punto importante cuando se trata de la corrupción cometida por fraudes es el uso de tecnología para evitar irregularidades o reducir su impacto. Los contratos son un tema complejo siempre que sean la principal forma en que los gobiernos transfieren dinero a otras organizaciones, incluidas las privadas. Este documento presenta una propuesta de estudio sobre el uso de la tecnología Smart Contracts (Contratos inteligente) en entornos Blockchain como una forma de enfrentar la corrupción en instancias gubernamentales. Los contratos inteligentes se pueden utilizar para todos los pagos gubernamentales como una forma de aumentar la transparencia de las transacciones, así como para evitar la sobrefacturación, siempre que los contratos y las licitaciones sean formas con características de cometer fraudes e irregularidades de dinero. Como investigaciones futuras, es importante verificar las barreras para la adopción de Blockchain, así como sus principales vulnerabilidades.
Nurul Hanim Razak, Haslenda Hashim, Nor Alafiza Yunus
et al.
Among many alternative fuels, oxygenated fuels like biodiesel and biomass-based energy (biofuel) such as bioalcohol have greater potential to enhance engine performance and mitigate particulate exhaust emissions in compression-ignition (CI) engines. The main objective of this study is to determine the optimal ternary green diesel (GD) blends formulation by identifying the most feasible diesel/biodiesel/alcohol that meeting the ASTM D975, Standard Specification for petro-diesel. Three steps of product design optimization (PDO) has been performed, (1) specify the fuel target properties based on Euro5; (2) optimize the formulation for ternary GD blends; (3) rank and select the optimal ternary GD blends. The ranking and selecting the optimal ternary GD blends were focused on the correlation of the higher cetane number (CN) over the cost of fuel. The PDO model indicated the most cost-effective and environmentally friendly diesel/biodiesel/alcohol ternary GD blends shall contain 74 % Malaysia petro-diesel, 16 % palm methyl ester (PME) and 1 % of butanol. Notably, the higher CN, the shorter the fuel ignition and the better the combustion efficiency. High CN fuels can significantly burn faster and more completely and hence reduce the harmful exhaust emissions such as SO2.
Chemical engineering, Computer engineering. Computer hardware
Craniofacial reconstruction studies the restoration of facial features according to the corresponding human skulls.It is widely used in many fields such as public security forensics,archaeology,medical plastic surgery etc.There are mainly two kinds of craniofacial reconstruction,the traditional craniofacial reconstruction and computer-assisted craniofacial reconstruction.The traditional manual craniofacial reconstruction is time-consuming and the results are often hard to achieve.In contrast,the computer-assisted craniofacial reconstruction is more realistic and efficient.This study focuses on the summarization and analysis of computer-assisted craniofacial reconstruction technologies,including knowledge-based craniofacial reconstruction method and statistical model method.We present elaborate introductions to some important methods,such as the craniofacial reconstruction method based on sparse soft tissue thickness,the craniofacial reconstruction method based on the thickness of dense soft tissue,the statistical deformation model method and the regression model method.The advantages and disadvantages of these methods are compared to help researchers get a comprehensive understanding on craniofacial reconstruction.
Pablo Etchepareborda, Francisco Veiras, Arturo Bianchetti
et al.
In this work we present an optical method for the direct determination of the piezoelectric coefficient of polymeric thin films. This is achieved through the measurement of nanometric mechanical displacements generated in the film when it is excited by low frequency harmonic electrical signals (0.5 Hz). The system is based on the temporal speckle pattern interferometry technique and the recovery of phase by using a bivariate empirical mode decomposition framework. The experimental scheme was used on a sample of vinylidene polyfluoride deposited on a glass substrate. The sample presents similar conditions to those found in the characterization of complex fluids by photoacoustic techniques. The measured value agrees with those obtained by other methods and with the value reported by the manufacturer.
Supplier evaluation is considered a key element in the procurement of resources. In this stage, a characterization of suppliers is carried out, based on a documentary review, interviews and experiences acquired in similar projects, which allows managers to make decisions about the project. Some of the main difficulties presented by methods in developing supplier evaluation are associated with inadequate uncertainty modeling and the lack of a mechanism for the treatment of multiple expert preferences on various criteria, which leads to loss of time and information. The general objective of this research is the application of fuzzy techniques for the evaluation of Video Surveillance technology suppliers, based on the fuzzy hierarchical analysis process and the 2-tuple linguistic representation model to treat uncertainty in decision making, based on the management of the information provided by multiple experts in their assessments. The obtained results can be easily interpreted by evaluators, without any loss of information.
In order to solve the problem of low recall rate in object detection with the deep reinforcement learning method,on the basis of simulating human visual mechanism,a dynamic searching hierarchical offset method is proposed.It uses the idea of anchors based on the original hierarchical searching method,which adds a region offset.This method avoids the limitations generated by hierarchical searching method,and makes the search more flexible.This paper combines the advantages of Double DQN and Dueling DQN,using Double Dueling DQN network structure as the deep reinforcement learning network of the agent.Experimental results show that the accuracy and recall ratio are higher than the original hierarchical searching method.