Computer Engineering Commons^™

Lightweight Assembly Workpiece Detection Algorithm Based On Improved Yolov8, Shuheng Wu, Yongkui Liu, Lin Zhang, Yingying Xiao, Lihui Wang

Journal of System Simulation

Abstract: To address the issues of low recognition accuracy and slow detection speed with existing deep learning-based object detection algorithms for robotic automatic assembly tasks, a lightweight assembly workpiece object detection algorithm based on YOLOv8 was proposed. The PConv was introduced to improve the C2f module, and a new Faster_C2f module was designed to enhance the detection speed of the model. The SIoU loss function was employed to optimize the location prediction accuracy of the CIoU loss function and improve the localization accuracy of small targets. The high-level screening-feature fusion pyramid networks (HS-FPN) structure was used to improve the Neck …

Automatic Guitar Transcription Of Polyphonic Music, Ritwik Patil

Master's Theses

Transcribing guitar music automatically is a complex task due to polyphonic overlap, tuning variations, and diverse playing techniques. Current transcription systems focus on identifying note pitches and timing while ignoring performance techniques that describe how the notes are played, treating guitar recordings as generic polyphonic audio and producing MIDI-like outputs that lose important information about articulation and style. To address these challenges, we propose an end-to-end transformer model for automatic guitar transcription. The system uses a T5-based encoder-decoder architecture that processes the Constant-Q Transform (CQT) of stereo audio input. The stereo representation helps separate individual guitar parts within a mix …

Green Ai-Enhanced Deep Learning Model For Breast Cancer Detection And Classification In Mammography Images: Bc-Net-512, Nesma Abd El-Mawla, Mohamed A. Berbar, Nawal A. El-Fishawy, Mohamed A. El-Rashidy

Mansoura Engineering Journal

This study champions a sustainable approach for developing a Deep Learning (DL) model for medical image analysis, specifically focusing on breast cancer (BC) detection in mammograms. By prioritizing low-computing algorithms to achieve high diagnostic accuracy while minimizing the model's environmental footprint, that aligns with the principles of Green AI. In this paper, an innovative architecture called BC-Net-512 was constructed for the classification of BC mammography. It is composed of lightweight Convolutional Neural Network (CNN) blocks for texture, density, and structure feature extraction and detection, a thin, fully connected layer for learning complex patterns and correlations in the extracted features, and …

3d Point Cloud Analysis With Classification, Segmentation And Few-Shot Learning, Jiajing Chen

Dissertations - ALL

Deep learning for 3D point cloud analysis has made significant progress, yet several critical challenges, including the following, remain underexplored: (1) traditional max-pooling operations discard a substantial portion of learned features, resulting in information loss and inefficient use of computational resources; (2) existing few-shot point cloud classification models lack robustness when faced with occlusion, missing points, and limited training data; (3) semantic segmentation methods often fail to fully exploit background–foreground interactions, leading to reduced accuracy. Moreover, in domains such as gait recognition and visual program synthesis, research has been largely dominated by 2D-based approaches, leaving the potential of point cloud …

Latent Action Trajectory Optimization, Rahul Milind Kandekar

Master's Theses

Learning from demonstrations offers a path to bypass the sample inefficiency of reinforcement learning, but obtaining action-labeled expert demonstrations remains expensive and often impractical. Learning from Observations (LFO) addresses this by learning policies from observation-only demonstrations. Recent LFO work relies heavily on behavior cloning: VPT and LAPO use observation-only data combined with limited action labels to train BC policies, while AIME offers an alternative policy inference approach but requires the majority of its training data to have action labels. Through systematic experiments in the Lunar Lander environment, we investigate whether latent action methods can function when state and action dimensionalities …

A Proposed Secure Video Using Lightweight Chacha20 Based On Chaotic Map, Hadeel Mohammed Taher, Ali Makki Sagheer

Iraqi Journal for Computer Science and Mathematics

Multimedia content has become a necessary part of our daily lives in many domains, including medical imaging, surveillance, and entertainment. Among all multimedia categories, video is mainly critical because of its high content density and large size. For that, ensuring the security of video has become a necessity for transmission and storage. This paper proposed an enhancement to the Chacha algorithm integration with a hybrid chaotic map for generating keys to secure video. The main motivation for the selection ChaCha algorithm is that it is fast, simple, and appropriate for widespread applications, whereas the motivation for using chaos is to …

Fasttree-Guided Genetic Algorithm For Credit Scoring Feature Selection, Rashed Bahlool, Nabil Hewahi Prof., Youssef Harrath Dr.

Research & Publications

Feature selection is pivotal in enhancing the efficiency of credit scoring predictions, where misclassifications are critical because they can result in financial losses for lenders and exclusion of eligible borrowers. While traditional feature selection methods can improve accuracy and class separation, they often struggle to maintain consistent performance aligned with institutional preferences across datasets of varying size and imbalance. This study introduces a FastTree-Guided Genetic Algorithm (FT-GA) that combines gradient-boosted learning with evolutionary optimization to prioritize class separability and minimize falserisk exposure. In contrast to traditional approaches, FT-GA provides fine-grained search guidance by acknowledging that false positives and false negatives …

A Comprehensive Review Of Dental Diseases Detection And Classification Based On Artificial Intelligence Techniques, Nermeen N. Noaman, Yasmin M. Alsakar, Naira E. Elazab, Waleed M. Mohamed, Mohamed E. Ezzat, Mohammed M. Elmogy

Mansoura Engineering Journal

In dentistry, many diseases, such as gum, cavities, and oral cancer, affect people of all ages. Early treatment and diagnosis are crucial for minimizing dental diseases' effect on overall health and saving money in the long run. Traditional dental diagnosis methods, such as manual probing and visual inspection, are time-consuming and can be subject to human errors. Hence, a computer-aided diagnosis system based on computer vision and artificial intelligence (AI) techniques is needed. The considerable progress in computer vision and AI techniques offers many possibilities in dental diagnosis based on dental X-ray imaging modalities. Dental X-rays are used to diagnose …

Making Robotic Reinforcement Learning More Efficient: Analyzing The Serl Framework, Faris Jugovic, Cleiver Ruiz-Martinez, Ryan Vander Stelt

Student Scholar Symposium

Teaching robots through reinforcement learning (RL) has made great progress, but real-world training is still difficult. Robots need lots of practice to learn, rewards can be hard to define, and resetting the environment after each attempt is often a challenge. The Sample-Efficient Robotic Reinforcement Learning (SERL) framework helps solve these issues by offering a ready-to-use, open-source software package that makes RL more practical for real-world robotics.

This project explores SERL and how it improves robotic RL by making learning faster and more efficient. SERL includes smarter ways to reuse training data, automatic methods for understanding rewards from images, and a …

Construction Of A Unified Knowledge Graph For Cyber Threat Intelligence, Moaz Usama Hassan Mr, Khaled , Nagaty, Noura Elmaghawry

Computer Networks

The rapid expansion and variety of cyber-threat information put enormous pressure on security operations centers (SOCs) that must convert unstructured data into understandable signals and make decisions upon it. This paper develops a Cyber-Threat-Intelligence (CTI) framework that integrates vulnerability information, product inventories, and weakness taxonomies into a domain-specific knowledge graph via automatic fusing. The proposed solution covers 284,296 CVEs, 101,644 CPE identifiers, and 965 CWE weaknesses, generating more than 800,000 typed edges linking threats, assets, tactics, and mitigations in an integrated CTI Knowledge graph. The graph was cross validated against four external standard datasets achieves full coverage of ATT&CK CAPEC, …

Implementing Controls To Mitigate The 2013 Target Data Breach: A Cost Benefit Analysis, Pricilia R. Fajong

Dissertations, Theses, and Projects

In 2013 Target had a data breach, which compromised 40 million credit/debit card accounts and 70 million customer records. The attackers exploited a vulnerability in a third-party vendor (Fazio Mechanical Services), to gain access to Target's systems. The breach cost Target over $250 Million (USD) in legal fees, investigation expenses, and reputational damage (Jones, 2025). Based on inflation rate, the 2013 Target data breach would cost over $340 Million (USD) today. In this study, a cost-benefit analysis was done to determine whether it would have been more cost-effective for Target to have invested in security controls rather than paying for …

Trustworthy Ai: Prohibited Practices, Ethical Principles, And The Identification Of Problems, Anna Karmańska

Journal of Global Awareness

The following considerations arise from the study of texts of documents of a legal nature and from the author’s judgments. They do not present the results of the author’s own empirical research; however, they constitute a factual study that is important for their undertaking in the next step. Having given concern but also hopes for AI, the author focuses her attention on issues that, not only in her opinion, have a strong bearing on the preservation of humanity in a digital environment and at the same time with technocratic features. These issues (prohibited practices, high-risk systems, and ethics) related to …

Object Tracking Based On Quantum Particle Swarm Optimization, Rajesh Misra, Kumar Ray

Journal of Global Awareness

In Computer Vision domain, moving Object Tracking is considered as one of the toughest problems. As there are so many factors associated like illumination of light, noise, occlusion, sudden start and stop of moving object, shading which makes tracking even harder problem not only for dynamic background but also for static background. In this paper we present a new object tracking algorithm based on Dominant points on tracked object using Quantum particle swarm optimization (QPSO) which is a new different version of PSO based on Quantum theory. The novelty in our approach is that it can be successfully applicable in …

Programing Ai With Ethics, Conor Anderson

Best Integrated Writing

As artificial intelligence grows increasingly ubiquitous, it’s pertinent to examine its fundamentals as well as its greater implications. Anderson discusses the ethical implications of AI.

View captioned video at https://youtu.be/LBTVEMm_C70

Leveraging High-Performance Cloud Computing To Model Underwater Acoustic Propagation And Scattering From Time-Evolving Rough Sea-Surfaces Using The Finite-Difference Time-Domain Method, James Alexander Higgins

Dissertations and Theses

This dissertation presents a two-dimensional (2D) Finite-difference Time-domain (FDTD) model for simulating underwater acoustic propagation and scattering from a one-dimensional (1D) time-evolving rough sea-surface. The techniques discussed are extendable to three spatial dimensions. Traditional acoustic modeling techniques often rely on a "frozen" sea-surface assumption, which proves inadequate for long-duration signals interacting with the time-evolving boundary at many different wave height displacements during its transit. To address this, a new FDTD update equation incorporating a variable subgrid is developed, significantly enhancing spatial accuracy at the boundary without increasing computational cost or compromising stability.

The model's accuracy is rigorously validated against established …

Ankimedbench: Evaluating Hierarchical Medical Knowledge In Language Model Embeddings, Neel Patel

UNLV Theses, Dissertations, Professional Papers, and Capstones

Despite achieving over 90% accuracy on medical benchmarks, recent studies show physicians cannot effectively leverage language models to improve clinical reasoning. Current benchmarks test isolated factual recall, but clinical practice requires hierarchical navigation through diagnostic categories—starting broad and narrowing systematically from chest pain to cardiovascular pathology to myocardial infarction to specific STEMI types. Existing evaluations cannot measure whether models preserve this taxonomic structure essential for clinical reasoning.

We introduce AnkiMedBench, built from 16,512 medical flashcards used by students preparing for licensing exams. Cards are organized across six hierarchy levels spanning 16 broad medical specialties to 672 specific diseases and conditions. …

Application Of Adversarial Volumetric Cnns To 3d Face Generation Using Latent Space Gaussian Embeddings, Ali Raad Abdulkareem, Marwa Jabberi, Islem Jarraya, Tarek M. Hamdani, Khmaies Ouahada, Adel M. Alimi

Iraqi Journal for Computer Science and Mathematics

Although 3D face generation is extensively studied in computer vision, most existing methods prioritize reconstructing 3D geometry from available 2D or 3D inputs rather than generating novel faces directly from latent representations. To bridge this gap, we present the application of Adversarial Volumetric Convolutional Neural Networks (AVCNN), a tailored adaptation of the vanilla 3D Generative Adversarial Network (3D-GAN), to 3D face generation using latent space Gaussian embeddings. We first assemble a custom 3D facial dataset to provide the requisite facial characteristics and to ensure sufficient coverage of geometric variation across identities. The generator, implemented as a decoder, maps latent space …

Retracted: Fuzzy Backstepping Approach To Stabilizing Fuzzy Parabolic Partial Differential Equations, Zainab John, Fadhel S. Fadhel, Samsul Ariffin Abdul Karim, Teh Yuan Ying

Iraqi Journal for Computer Science and Mathematics

In this work,\floatquery[-14pc]AU: Please provide ORCID ID for remaining authors. we study the solvability and stability of the fuzzy reaction-diffusion equation and the fuzzy diffusion equation with nonhomogeneous boundary conditions. Firstly, we derive the fuzzy exact solutions by converting the nonhomogeneous boundary condition to a homogeneous condition, then we prove the instability of the solution by using simulation approach with the help of Maple program. From this, we have decided to implement the fuzzy backstepping approach to developed the stabilization of the fuzzy parabolic partial differential equations. By using this approach together with the generalized Hukuhara (gH) derivative, we were …

Dynamic Admittance Parameterization For Non-Prehensile Multi-Robot Transport With Optimal Coordinated Planning, Calvin J. Stahoviak

Computer Science ETDs

The Dynamic Admittance Parameterization of Non-Prehensile Multi-Robot Trans- port with Optimal Coordinated Planning (DYNAMO) architecture offers a practical framework for cooperative payload transportation using two robots equipped with nonholonomic mobile bases and four-degree-of-freedom manipulators. Coordinated mobile manipulation is a difficult problem in robotics, and the non-prehensile case is even more challenging than its prehensile counterpart because the robot bases and the payload are dynamically coupled. DYNAMO adapts arm motion in response to interaction forces and generates coordinated base trajectories that account for this coupling. Robust payload transport is achieved through the combination of opti- mal planning and adaptive compliant control, …

Parameter-Efficient Multimodal Adaptation: Ocr-Integrated Lora For Textvqa And Captioning, Karthik Ganesh Malini

Master's Theses

Vision-Language Models (VLMs) have emerged as transformative technologies for multimodal AI, yet they face significant hurdles in processing text-rich images required for enterprise applications like document understanding, medical imaging, and industrial inspection. Current VLMs struggle with accurate text extraction and reasoning, often exhibiting high hallucination rates and poor Optical Character Recognition (OCR) token utilization. To address these limitations, this research presents a comprehensive framework for optimizing parameter-efficient Low-Rank Adaptation (LoRA) fine-tuning strategies on state-of-the-art architectures, including LLaVA-1.5 and BLIVA-FlanT5. Our methodology integrates enhanced OCR token utilization, faithful caption generation, and specific hallucination mitigation techniques. We employ a multi-dimensional evaluation protocol …

Technological Disruption And Regulatory Response: The Case Of Decentralised Finance, Jakub Wisła, Jolanta Bartoszewska

Journal of Banking and Financial Economics

This article examines responses to the regulatory challenges posed by decentralised finance (DeFi), a fast-evolving domain of blockchain-based financial innovation. It investigates the factors shaping divergent regulatory strategies, with a focus on the European Union’s comprehensive cryptoasset framework and selected comparative insights. Adopting a qualitative legal methodology – combining doctrinal-functional analysis, multivocal literature review, and two case studies – the authors explore how regulatory responses are influenced by three key variables: legal tradition, the financial function performed by blockchain-based solutions, and the level of technological and institutional autonomy. The case studies – Bitcoin as a payment instrument and cryptoassets as …

Leveraging Google Earth Engine For Computationally Efficient Pixel-Level Analysis And Vector Delineation From Satellite Data., Rishita Garg

Theses and Dissertations

Analyzing large-scale, high-resolution satellite imagery is a computationally intensive task requiring time and computing resources. This can be accelerated using cloud computing platforms such as Google Earth Engine (GEE) where computational and storage requirements can be scaled based on demand. However, cloud-based platforms for processing high-resolution imagery remain underutilized in environmental applications such as agriculture, and forest health. This thesis explored the application of GEE to two geospatial problems in agricultural conservation and disease mapping in forestry: 1) Extraction of agricultural field boundaries from Sentinel-2 satellite imagery, for use in conservation, precision agriculture, land management, and organization, etc., and 2) …

Conditional Generative Adversarial Network Framework For Iot Anomaly Detection, Henry Onyeka

Tennessee State University Alumni Theses and Dissertations

The growing scale and complexity of Internet-of-Things (IoT) edge networks complicate anomaly detection, particularly in identifying sophisticated Distributed Denial of Service (DDoS) attacks and zero-day behaviors under highly dynamic and imbalanced traffic conditions. This thesis proposes SD-CGAN, a Conditional Generative Adverserial Network optimzied with Sinkhorn Divergence as a geometry-aware one-class framework for robust IoT anomaly detection. SD-CGAN trains solely on benign traffic flows to learn a stable representation of normal traffic. To address class imbalance and improve the variety of the sample, we combine SD-CGAN with CTGAN-based synthetic data augmentation. Replacing the adversarial objective function with Sinkhorn Divergence yields smooth …

Improving Road Safety Through Multimodal Deep Learning For Driver Drowsiness Detection, Hadel A. Hussain, Mohammed A. Subhi, Ahmed S. Al Tmeme, Ahmed D. Radhi, Marwan Ali Albahar

Iraqi Journal for Computer Science and Mathematics

One of the most common causes of road accidents globally is driver drowsiness and it needs solutions that are reliable and can be applicable in numerous real-life situations. We present this paper with the aim of developing a deep-learning system that is capable of reliably detecting drowsiness in diverse and varied conditions across different drivers, environments, and sensor types. Our system is known as Multimodal Attention Network (MMAN), which combines information of eye and head movement, heart-rate and breathing pattern, and vehicle-dynamics signal. MMAN has a gradient-reversal layer that enables the layer to be domain-adaptive such that it does not …

Deep Learning-Based Fog-Cloud Approach Intrusion Detection System In Iomt, Yahya Rbah, Mohammed Mahfoudi, Mohammed Fattah, Younes Balboul, Said Mazer, Moulhime Elbekkali

Iraqi Journal for Computer Science and Mathematics

The Internet of Medical Things (IoMT) creates an interconnected environment linking humans, devices, sensors, and systems, enhancing healthcare services through advanced technologies. Nonetheless, these IoMT devices are susceptible to cyberattacks, which can endanger patient safety and healthcare services. To identify and mitigate cyberattacks in IoMT, techniques such as threat intelligence, log monitoring, and intrusion detection systems are employed. As attackers evolve their strategies, there is a growing trend towards leveraging artificial intelligence to achieve more predictive and accurate attack detection. Since IoMT devices are inherently low-power, they require minimal computing resources. Existing intrusion detection systems are generally trained in the …

Multitaskvenationnet: A Multi-Task Deep Neural Network With Strip Pooling And Hybrid Upsampling For Leaf Vein Segmentation, Ishak Ariawan, Ahmad Ashari, Moh. Edi Wibowo

Iraqi Journal for Computer Science and Mathematics

Leaf vein segmentation is a critical task in plant phenotyping and species classification, yet it remains challenging due to the hierarchical, curvilinear nature of veins and interference from complex backgrounds. Existing methods face three key limitations. First, they lack directional context modeling, leading to blurred vein boundaries and the omission of fine venation. Second, they fail to effectively capture global dependencies, limiting semantic coherence across spatial regions. Third, they do not incorporate explicit mechanisms for detecting vein discontinuities, which is essential for complete topological understanding. To address these challenges, we propose MultiTaskVenationNet (MTV-Net), a multi-task deep segmentation framework that integrates …

Exploring Interactive Robotic Music Therapy Systems For Rehabilitation: A Survey Paper, Hector A. Salinas Gordillo

Discovery Undergraduate Interdisciplinary Research Internship

Interactive robotic music therapy introduces an innovative opportunity, where human guided musical interaction with robotic systems can create adaptive and engaging therapeutic experiences. This survey explores the current state of research at the intersection of robotics, music, and rehabilitation, focusing on emerging technologies such as human robot interaction methods and system designs that help enable real time, interactive music therapy.

Potential patient groups include individuals undergoing motor or cognitive rehabilitation, such as those recovering from stroke, living with Parkinson’s disease, cerebral palsy, or other motor impairments, as well as individuals with developmental disorders or limited mobility.

Traditional rehabilitation exercises may …

Spike Timing Depended Plasticity Produces Unsupervised Learning Of Synergistic Muscle Feedback In A Synthetical Neural Network, Mark Allen Pupkiewicz

Dissertations and Theses

This study investigates how type Ia feedback from muscle spindles can be organized into groups representing agonistic muscle pairs through Spike Timing Dependent Plasticity (STDP). A single degree of freedom joint is actuated with four biologically modeled muscles forming two agonistic pairs. In order to emulate the sensory dynamics of biological muscle spindles, sensors in the model record the active length and velocity states of each muscle, the two primary factors eliciting type Ia afferent responses. In biological networks, synapses from Ia sensory neurons frequently activate interneurons representing agonistic muscle sources. This research investigates whether this organization can emerge in …

A Predictive Model For Multi- Week Respiratory Risk From Red Tide On Florida’S Gulf Coast., Elmer S. Ochaeta

Computer Science and Engineering Faculty Publications

Florida’s Gulf Coast red tide (Karenia brevis) can put toxins into the air, making people cough, irritating the throat, and worsening asthma or other breathing problems especially when winds blow from the ocean toward the beach. Right now, most public updates don’t really help with the question people actually ask when planning a weekend or vacation: “Will going to or close to the beach be risky in the next few weeks?”.

In this project, I build a weekly early warning system that estimates respiratory risk for specific beaches and predicts that risk 2 to 4 weeks ahead. The study covers …

3d Printed Portable Automatic Pill Dispenser​, Amber M. Ocasio

Publications and Research

Medication adherence is a major public health concern, particularly among patients with chronic illnesses. Reports from the National Institutes of Health indicate that adherence rates are significantly lower for chronic conditions, with patients taking only ~50% of medications prescribed. Unintentional non-adherence—such as forgetting doses—is more prevalent (62.9%, 47.1%, 46.9%) than intentional non-adherence, and the consequences include medication waste, disease progression, reduced functional abilities, lower quality of life, and increased reliance on medical resources. Because existing automatic pill dispensers cost over $100 on average, they remain inaccessible for many lower-income patients who could benefit from such technology. This project addresses this …