International Conference on Cloud Computing and Computer Networks

CCN-51 cocoa, one of the two main varieties exported worldwide by Ecuador, due to the lack of technology and poor agronomic practices, is constantly attacked by a number of pests that affect its production, affecting the growth stages of the plant. Another factor damaging the plant is the frequent climate changes, mainly due to excessive rainfall increasing humidity levels. These conditions damage the flowering and fruit set, leading to Moniliasis as one of the primary diseases. Given that the crops are located far from urban areas, conducting analyses is time-consuming and costly. Consequently, many producers resort to excessive chemical use to manage pests and diseases. Where, this research project is proposed, consisting of developing a mobile application that by scanning images in a controlled environment allows the detection of diseases in the CCN-51 cocoa fruit. The mobile application will use its camera to scan the fruit and, using a trained image recognition model, predict a diagnosis of the disease present in the cocoa fruit.

In recent years, with the deepening of the exploitation of marine resources, the demand for marine environment exploration is also increasing. However, due to the complex and changeable marine environment, the exploration risk of underwater manned spacecraft is relatively high. Autonomous underwater vehicle (AUV) has become an essential tool to explore and exploit marine resources. During the execution of underwater tasks, AUVs often encounter collision and entanglement of various obstacles, which will pose a fatal threat to AUVs. Forward-looking sonar (FLS) is one of the main sensors used by AUVs to detect underwater targets. By continuously transmitting sound waves and receiving echo signals, AUVs equipped with multi-beam forward-looking sonar can generate real-time visual field acoustic images, and then detect the images with target detection algorithms to solve the above problems. However, the visual acoustic image has the characteristics of more texture feature information and less semantic feature information, and it requires real-time reasoning. Some existing classical target detection algorithms are mostly designed for some data scenes with strong semantic meaning, and cannot be well adapted to the visual acoustic image. Therefore, a new target detection algorithm, Swin_FLS, is proposed in this chapter. The improved Swin_FLS can more fully extract the texture features in the forward-looking sonar image and adapt to the characteristics of the forward-looking sonar image data set. Finally, 85.2 mAP and 14.1 FPS are obtained in the test set. It surpasses the accuracy of some classical algorithms directly applied to the forward-looking sonar image data set.

Efficient facial landmark detection has been applied to various fields, such as driverless driving, facial beautification technology, facial expression analysis, etc. However, in specific practical tasks, there are still some situations where facial expression cannot be correctly recognized or analyzed. This paper proposes an improved MobileNetV2_re method to improve the loss of accuracy of key points of the problem of pixel-in-pixel Net (PIPNet) in the existing facial landmark detection task. We use the ghost module to replace part of the inverted residual block from the original model, build a new MobileNetV2_re network, and improve the accuracy of the model. It is proved that the situation where high NME and low AUC of PIPNet in the original network MobileNetV2 can be effectively improved by comparing the tested normalized mean error (NME) and the area under the curve (AUC) value and selecting a better network. Compared with MobileNetV2, Resnet18, Resnet50, and Resnet101, NME of MobileNetV2_re in PIPNET is reduced by about 14.07%, AUC of MobileNetV2_re in PIPNET is increased by about 7.52%, and it shows higher accuracy in efficient facial landmark detection.

This chapter presents a framework of nonlinear filter-based compressed sensing (CS) applied on sparse view CT image reconstruction. The conventional algorithms of quadratic-form regularization have been recognized that they fail to take the discontinuities of images into account. They may lead to over smoothed of object boundaries or fine structures in the reconstructed images. The proposed method considers sequential changes of all the local pixels in a designed window. A kind of nonlinear filter, which is median filter, bilateral filter, or non-local weighted means filter, was embedded in the CS framework, respectively. The image reconstruction problem can be treated as a cost function minimization problem. The proposed cost function consists of data fidelity term and penalty term. The penalty term, in which the nonlinear filter works, was designed and tested in ℓ0 norm, ℓ1 norm, and ℓ2 norm. Proximal splitting theory was used for constructing iterative reconstruction methods. Furthermore, we carried out calculation acceleration by constructing row-action structure. In this chapter, we applied the new method to practical medial CT images (dental, chest, and cranial images), and it showed a superior effect in image smoothing, object boundary extracting, and texture preserving. Appropriate adjustment of nonlinear filter parameters is a little complicated, and we will promote the automatic parameter setting or reducing parameter numbers in our future work.

Despite the advancements in security, threats have become more sophisticated than ever-leading companies to think outside the box. Cyberattacks are becoming increasingly sophisticated due to hybrid work. Business continuity often took precedence over security concerns as organizations scrambled to comply with shifting regulations. Now that more people are working remotely via cloud services, IT needs assistance from cybersecurity experts. There are several sources that can pose a threat, including office printers. It is not uncommon for printers to have hundreds of potential entry points for hackers, who can then bring a system to its knees by taking control of one of the printers attached to it. In today’s world, printers are very much computers and are often connected to the Internet. Having advanced abilities makes it easy for cybercriminals to access them. This paper analyzes printer attacks from the past and provides a general methodology for analyzing printer security. Our methodology will be used to conduct online surveys of experienced IT practitioners to explore their exposure to social engineering attacks and security concerns related to printers, digital copiers, and 3D printers. Passive reconnaissance will be conducted to determine the extent to which some network protocols are exposed by these devices. A compiled checklist has been consolidated to be considered by businesses as a risk mitigation technique to secure the devices from vulnerabilities and attacks.

As a consequence of the exponential growth of current technology and, by nature, information digitization, there is an increasing number of final devices on the Internet of Things (IoT). To address these requirements, it is appropriate to access new capabilities of information technologies (IT) and operational technologies (OT) and have the potentialities offered by the Cloud, but from an environment close to the production plant, which implies a reduction in broadband costs and low dependence on latency. With this approach, this study presents a Fog Computing (FC) alternative implemented for object recognition through Artificial Intelligence (AI). The architecture proposed consists of deep neural networks (DNNs) processing nodes, the main Fog node for Docker-based provisioning and Kubernetes orchestration; in addition, Prometheus was used for collecting dynamic metrics, such as central processing unit (CPU), random access memory (RAM) and power absorbed, and afterward, Grafana was used for the analysis and visualization of their trends. After experimentation, the Fog architecture was obtained by balancing the workload in the two processing nodes, achieving an improvement of more than 50% in the provisioning and orchestration processing time compared to an architecture constituted by a single node.

In the current age, ensuring the security and confidentiality of information transmitted via the Internet is a crucial issue for both individuals and companies, and plenty of technology companies and their technologies are offering such features. Like other numerous technologies, virtual private networks (VPNs) technologies have been applied to the modern world for security. Different VPN companies that offer free to paid services help provide security capabilities that a user would need to use when facing a vulnerability in using VPNs. These offers could include encryption, firewalls, and encryption with technologies such as tunneling, portal, and remote desktop architectures. This research has gathered information regarding user interest in VPNs, as well as seeking their confidence and knowledge regarding the security and privacy of different VPN clients. The results of the survey show that users are knowledgeable of VPNs, their services, and their acknowledgment of their vulnerability that hackers could use to their advantage.

To solve the problem that the convergence speed of Grey Wolf Optimization (GWO) is not fast enough and tends to fall into local optimization, the improved Grey Wolf Optimization based on logarithmic inertia weight (LGWO) is proposed. LGWO utilizes the characteristics of logarithmic function to realize the nonlinear adjustment of inertia weight, thus better balancing the global exploration and local mining capabilities of the GWO. Meanwhile, the logarithmic inertia weight strategy is introduced into the location update of the GWO to deal with the location update process of grey wolves, which reduces the possibility of the algorithm falling into local convergence and accelerates the convergence speed. Five classical test functions are used to test the optimization performance of LGWO. Compared with the existing swarm intelligence algorithm, LGWO accelerates the convergence speed and improves the convergence accuracy and stability of the GWO.

Radio frequency identification (RFID) is a technology that uses radio waves to communicate between a reader and a tag attached to an object, to identify and track it. RFID systems consist of a reader, an antenna, and a tag or transponder (Ait et al., Math Model Comput 8(4): 616–626, 2021). In this research, the aim was to examine the vulnerabilities of RFID-related physical controls in an infrastructure, to identify potential vulnerabilities, and to assess the associated risks. The findings of this study provided important insights into the current state of RFID-related physical controls in an infrastructure and will assist organizations in improving the security of their RFID systems.

This essay proposes two methods to analyze bee populations in a given period. The first method is a quantitative analysis of the correlation between time and population, establishing a time–population model for bees. However, this method fails to provide a precise enough result. For improvement, the analysis of bee populations is augmented with more comprehensive factors (both positive and negative), creating a unified measure to calculate the total change in population percentage by assigning weights to each individual factor. During the construction of these two methods, we completed the following five steps: Find relevant data with a numerical correlation between time and population: Data containing relevant information like time and population were downloaded from credible sources. Then, the data were fitted with linear regression to reveal the relationship between the population and time. Find possible factors that affect bee populations: External and internal factors were identified through a literature review of research articles and reputable online sources. Among these, five factors were deemed the most critical and to be used in this chapter later. Assign weights to each factor through the Entropy Weight Method (EWM) and Analytic Hierarchy Process (AHP): With EWM or AHP, a different set of weights was assigned to the factors. However, in this paper, neither of these two was used alone. Instead, a unified model that learns from both methods and hence generates a better weight for each factor is proposed and explained. Analysis of beehives needed to pollinate a 20-acre area: Parameters for the model were identified, defined, and populated using relevant data. Finally, the minimum and the maximum number of beehives that satisfy the requirements were calculated and an average of the values was obtained. Testing of the model on Buhlmann 1985: With the fully calculated weights of different factors through the integrated method, the model was tested to see if the weight assignments were reasonable. To do this, the result obtained from this model is compared with data approached by Buhlmann (1985) as an evaluation of this model.

The capabilities of artificial intelligence (AI) and deep Learning are increasing rapidly with the increasing computing power and specialized microprocessors. A very interesting architecture, generative adversarial networks (GANs), is at the forefront of innovation. Some examples of what GAN networks are used for are text-to-image translation, image editing/manipulation, recreating images of higher resolution, and creating three-dimensional objects. When it comes to audio, Google WaveNET, Parallel WaveNET, and its successor Tacotron 1 and 2 are the frameworks of choice to create synthetic-based audio. In cases where there is not enough training data, one can synthetically generate data for further research and training. Methodology wise qualitative data samples can be synthetically generated for any language. This paper showcases data generation for the Afrikaans language. Here, we used a trained network to create Afrikaans speech clips based on text. Finally, when generating the same sentence multiple times, the clips have different emotional states. These clips are then verified, categorized, and used to train another network.

The decline of the bee population poses threats to the production of considerable types of crops that require pollination. The prediction of the bee’s future population has therefore become a valuable research topic. For Problem one, we tried to solve it in mainly two ways: using the Grey Forecast Model and using differential equations. For data that were missing, we processed them by normalization at first and then regressed to find the abnormal data, and filled the missing data with average data after deleting abnormal data. For the Grey forecast, we use three types of models and compared their respective results with true values to pick the one with the most accurate output and use it to predict the population of bees. For the differential equation method, we simply express the rate of increase in population in terms of several variables (in the differential equation) and solve the equation to obtain the future population. For Problem two, we do a sensitivity test on the bee population. We applied the Random Forest model here to determine the importance of each variable. During the evaluation of the model, we test four sets of data and compare the Random Forest results with the true value. It turned out to be that the final model predicts the population precisely, which has proven that it is reliable. At last, we change the sensitivity of each variable for a 100% change and tell the importance of the variables. For Problem three, we get the model of the possibility of a plant being visited by a bee in a beehive system at any distance, and then we use this matrix to simulate the area and calculate the possibility at any point. After determining a possible lower bound, we can get the area that can reach the bound which is the area the current beehive system can serve. By changing the number and the positions of beehives, we can get the maximum area the system can serve at any time. We can also calculate the possibility considering the planting density and the population of bees so it can be related to problem 1.