Precision agriculture brings together most innovative topics from research in the fields of vision, artificial intelligence and robotics aimed at developing sustainable and resilient solutions. In this context, crop identification is a crucial issue. This paper exposes a new algorithm to enhance real-time crop/weed instance segmentation, merging a learning-based method for instance segmentation and a feature model-based image processing strategy that relies on the vegetation characteristics of the crops. The proposed algorithm compensates for the shortcomings of each method performing independently. That is, the image processing strategy fulfils crop segmentation by generating finely refined masks; however, some weeds are segmented incorrectly. On the other hand, instance segmentation methods perform well on weeds and crops but, when inaccurate bounding boxes are present, imperfect masks are generated. The experiments are conducted in different evaluation campaigns within the ACRE international competition framework.

This study compares novel and established feature extraction techniques for classifying leaf diseases. Conventional methods like Transfer Learning and DCT are contrasted with novel techniques like Segmentation using a hetero- geneous data set of healthy and diseased leaf images. With little training data, transfer learning is beneficial, but with large data, more efficient al- gorithms may be more suitable, such as the custom segmentation model described in this work. The development of efficient plant disease detection systems using the information from these discoveries would improve agricul- tural practises and increase food security.

Circuit theory cuts across every part of electrical engineering and is a fundamental skill for electrical engineers. However the abstractions involved can present challenges for students. In this work, a problem solving approach was employed to help students motivation and interest in class attendance.  This study looks at performances in the final examination and analyses them with a view to developing better approaches and tools for future delivery of such a course.

In the ongoing transition towards a sustainable energy system, the electric power system increases in complexity and must adapt to rapid changes and new uncertainties. Thus, development and application of appropriate probabilistic methods is all the more important. Reliability analysis based on Monte Carlo simulation (MCS) has become increasingly popular, and its strength lies in the ability to account for a large number of random variables, general stochastic processes and assessing the probability distribution of the output. However, power system reliability is governed by relatively rare interruption events which poses a fundamental challenge to MCS. This paper presents a variance reduction technique for Monte Carlo based reliability analysis which combines resampling and the cross entropy (CE) method. The motivation for the work is to mitigate the computational burden related to rare event sampling, while at the same time preserving the flexibility of MCS by introducing few assumptions on the stochastic model. The method is demonstrated on a synthetic test system and gives a speedup of about 10 times compared to a crude simulation.

Alzheimer’s Disease has been a long standing mental disorder that we have gained a lot of knowledge about. However, no matter how much we have learned, we still are yet to discover a magnitude more about the disease. Alzheimer’s disease is a progressive neurological disorder characterized by cognitive decline, memory loss, and behavioral changes. It is the most common cause of dementia, impacting a person’s ability to perform daily tasks and ultimately affecting their quality of life. We can study this using brainwaves. The 5 types of brainwaves each correlate to a different type of mental activity, which gives more information on various diseases like Alzheimer’s. Through this paper, we hope to find common trends on the brainwaves of an Alzheimer’s patient compared to a control subject with no mental disorders. All patterns will be tried to conclude to the reason behind that data. The Muse 2 Headband was used to collect the recordings of brainwave data in this study. Recordings of two subjects (one of which is the control) in two different conditions (active and resting) will be tested. The EEG data will then be compiled by MATLAB and put on a Spectral Plot to analyze the Electrical Brain Responses. The results show that readings in all brain waves other than Beta and Gamma were inactive and similar between both subjects. Beta and Gamma wave analysis proves that the Alzheimer’s Disease subject had lower activations in their brain due to a higher log of frequency power in those waves.

This paper introduces an AI-driven language model (LLM) designed to facilitate users’ comprehension and adept usage of different NVIDIA SDKs (Software Development Kits) and toolkits. The central goal is to provide a LLM which establishes an interactive and intuitive platform that offers inclusive insights, practical illustrations, and expert direction on NVIDIA’s diverse SDKs and toolkits. Harnessing the capabilities of language models and NVIDIA’s toolkits, this study strives to streamline devel- opers’ learning process, enabling them to harness NVIDIA’s technological advancements with enhanced proficiency.

In recent years, there has been a significant increase in vehicles, resulting in areas that are unable to supply online connectivity, due to the lack of infrastructure development, and high construction costs. This, in turn, can cause vehicle outages. Therefore, this paper proposes a method to improve coverage using V2V communication and designing promotional contracts for relay vehicles. Relay vehicles are rewarded according to the contract, and the aim of this system model is to select the appropriate relay, allocate power, determine the optimal amount of data to transmit, evaluate of the relay reward, and ultimately maximize the utility of the base station. The paper considers halfduplex (HD) and full-duplex (FD) relays separately. However, optimization problems are NP-hard problems, and to solve this issue, the paper uses successive convex approximation and Taylor expansion to approximate the problem. Additionally, the paper presents an algorithm for relay selection using a simulated annealing algorithm. The simulation results examine the amount of data provided by two models, rewards and objective functions, and the results show that the FD model outperforms the HD one. Furthermore, the system is extended and solved for multi target and multi relay vehicles.

A document by Mohammad Abdel-Rahman . Click on the document to view its contents.

The integration of blockchain technology in financial sectors has emerged as a topic of burgeoning interest, with credit business being a pivotal area of exploration. This study aims to investigate the impact of blockchain technology on the operational efficiency and security of credit business processes. Employing a mixed-methods research design, data was collected from 15 credit institutions that have adopted blockchain technology, alongside a control group of 15 credit institutions operating on traditional digital platforms. Key performance indicators (KPIs) such as transaction speed, error rate, fraud incidence, and operational costs were analyzed. The findings reveal a significant enhancement in operational efficiency and security in credit institutions leveraging blockchain technology. Blockchain adoption was associated with a 70 % reduction in fraudulent activities, a 50 % increase in transaction speed, and a 40 % reduction in operational costs. Furthermore, the decentralized nature of blockchain significantly enhanced transparency and trust among stakeholders, fostering a more robust and resilient credit business ecosystem. The study underscores the transformative potential of blockchain technology in redefining the operational paradigms of credit businesses, thereby contributing to the broader discourse on blockchain’s applicability in financial sectors. Future research is recommended to explore the long-term sustainability and regulatory implications of blockchain integration in credit business operations.

Artificial Intelligence (AI) Foundation Models (FMs), pre-trained on massive datasets, have recently emerged as a pivotal asset in a wide array of tasks. Examples of FMs include Large Language Models (LLMs), Large Vision Models (LVMs), and Large Multimodal Models (LMMs). The adaptability of FMs, achieved through finetuning, enables these models to perform exceptionally across diverse domains. However, the fine-tuning process often entails data centralization, which raises privacy concerns. For instance, in healthcare, hospitals might want to fine-tune an AI model on patient records. Sending this data to a central server for fine-tuning can raise privacy concerns. To mitigate the privacy challenges, this research seeks to employ privacy-preserving technologies such as federated learning (FL), differential privacy (DP), and emulator-based tuning (i.e., offsite tuning) in combination with parameter-efficient fine-tuning (PEFT) techniques to refine FMs without compromising data privacy.

Most existing methods for Magnetic Resonance Imaging (MRI) reconstruction with deep learning use fully supervised training, which assumes that a high signal-to-noise ratio (SNR), fully sampled dataset is available for training. In many circumstances, however, such a dataset is highly impractical or even technically infeasible to acquire. Recently, a number of self-supervised methods for MR reconstruction have been proposed, which use sub-sampled data only. However, the majority of such methods, such as Self-Supervised Learning via Data Undersampling (SSDU), are susceptible to reconstruction errors arising from noise in the measured data. In response, we propose Robust SSDU, which provably recovers clean images from noisy, sub-sampled training data. Robust SSDU trains the reconstruction network to map from a further noisy and sub-sampled  version of the data to the original, singly noisy and sub-sampled data, and applies a Noisier2Noise correction term at inference. We also present a related method, Noiser2Full, that recovers clean images when noisy, fully sampled data is available for training. Both proposed methods are applicable to any network architecture, straight-forward to implement and have similar computational cost to standard training.  We evaluate our methods on the multi-coil fastMRI brain dataset with a novel denoising-specific architecture and find that it performs competitively with a benchmark trained on clean, fully sampled data.

Quantum Computing is a field that is gaining recognition at a rapid rate around the world. This review shall address the unique elements of quantum computing with respect to classical computing, the current situation of quantum computing (i.e. Noisy Intermediate Scale Quantum or NISQ era computers) and the potential of such technology for the future. The review further seeks to discuss the problems that need to be overcome to attain quantum computing feasibility and the use of hybrid computing as an alternative that can derive from the benefits of the quantum world.

This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible.

This article aims to ensure continuity between classical methods of ultraviolet microscopy or/and micromanipulation using ultraviolet microbeam and lens-less ultraviolet microscopy and microbeam exposure of cells and tissues. Considering the history of the development of the method and the possibility of working with different methods of mechanical scanning, the authors propose to use mirror galvanometers and an electromechanical scanning system in the mechanical engineering of lensless microbeam installations. These installations make it possible to provide both scanning with an ultraviolet microbeam to obtain a line scan image, and precision micromanipulation at the level of individual cells. This article is a preprint (or reprint) associated with the authors’ presentation at the September 2018 seminar dedicated to the memory of Sergei Chakhotin, the developer of the first UV microbeam installations. The deposit of this article in the technical archive in September of this year is associated with a memorable date associated with the dates of the life of Sergei Chakhotin (13 September 1883 – 24 December 1973). The first article by one of the authors, which formed the basis for this presentation, was also timed to coincide with the memorable date of Sergei Chakhotin and was presented at the Christmas seminar at the end of 2013 (December 24), and was published in an issue of the Ukrainian bioengineering journal in 2014.

In the realm of linear algebra, the notion of matrix inversion plays a crucial role. While the inversion of square matrices is well-known and results in a unique inverse, however, the non-square inverse matrice is not unique and in fact, the number of inverses for a non-square matrix can be as vast as q^m(n−m), where q signifies the order of the underlying field. In this paper, we embark on a journey to construct these elusive inverse matrices, harnessing the power of arbitrary fields. Arbitrary fields, including prime fields, finite fields, real fields, and complex fields. These fields find practical applications that are essential to contemporary technology. I have written 5 MATLAB programs that able to construct random inverses in different fields based on the given algorithm.

This paper investigates the discussion on autonomous weapons, delving into the legal, ethical and practical implications of AI and military technology. We find that as the deployment of many autonomous weapons becomes increasingly prevalent, concerns regarding accountability, transparency, and compliance with international humanitarian law have escalated, leading to the question of human control. By utilizing a multi-dimensional research approach through policy reviews, case studies, and ethical examinations, this study emphasizes the necessity of establishing clear ethical frameworks, and legal regulations to guide the development, deployment, and use of autonomous weapons. The findings of this research contribute significantly to the field of AI ethics, military strategy, and international law. The key findings highlight the potential risks and ethical challenges associated with the lack of human control, providing valuable insights to policymakers, researchers, and defense experts to promote responsible AI governance and safeguard human rights in the realm of military technology.

The Riemann Hypothesis, a cornerstone in number theory, illuminates prime number distribution’s average and the deviations from this mathematical norm. Originating in Riemann’s seminal 1859 paper, it posits that the elusive zeros of the zeta function inhabit the complex plane with a real part fixed at 1/2. This hypothesis offers a one-million-dollar challenge for those capable of providing a valid proof in a respected mathematical journal. Remarkably, a parallel emerges when considering the shared challenges between the Riemann Hypothesis and the domain of medical predictions, where accurate outcome forecasting remains a formidable task in intricate scenarios. Stephen Wolfram’s Principle of Computational Equivalence underscores the computational essence of complex processes. Within this concise communication, we delve into two conjectures related to the generalized Riemann Hypothesis for Dirichlet L-functions, transcending disciplinary boundaries to intertwine elements of number theory, medical science, and captivating references such as Abraham Lincoln and a hypothetical set of dogs with cardinality one.

Efficiently predicting drug response remains a challenge in the realm of drug discovery. By leveraging two types of multiomics data, transcriptomics, and genomics, we create a comprehensive representation of target cells and enable drug response prediction for personalized medicine. To address this issue, we propose four novel model architectures that combine graph transformers with varying positions of multiheaded self-attention mechanisms. A majority of our architectures utilize multiple transformer models, one with a graph attention mechanism and the other with a multiheaded self-attention mechanism, to generate latent representations of both the drug and omics data, respectively. Unlike previous approaches that apply attention only to one data type, either drug or genomics, our model architectures employ this technique for both, with a goal to procure more comprehensive latent representations. The latent representations are then concatenated and input into a fully connected network to predict the IC-50 score, a measure of cell drug response. We experiment with all four of these architectures and extract results from them. The novel model without the multiheaded self-attention mechanism seems to give us the most accurate results on our holdout set. Our study greatly contributes to the future of drug discovery and precision medicine by looking to optimize the time and accuracy of drug response prediction, as well as using multiomics data for a personalized approach to the problem.