Engineering Commons^™

A Memory Contention Responsive Hash Join Algorithm Design And Implementation On Apache Asterixdb, Giulliano Silva Zanotti Siviero

Computer Science and Engineering Master's Theses

Efficient data management is crucial in complex computer systems, and Database Management Systems (DBMS) are indispensable for handling and processing large datasets. In DBMSs that concurrently execute multiple queries, adapting to varying workloads is desirable. Yet, predicting the fluctuating quantity and size of queries in such environments proves challenging. Over-allocating resources to a single query can impede the execution of future queries while under-allocating resources to a query expecting increased workload can lead to significant processing delays. Moreover, join operations place substantial demands on memory. This resource’s availability fluctuates as queries enter and exit the DBMS. The development of join …

Sovia: Sonification Of Visual Interactive Art, Lauryn Gayhardt

Computer Science and Engineering Master's Theses

Therapeutic Computational Creativity is an emerging domain that challenges us to explore applications of Computational Creativity systems to mental health and wellness. This work presents SOVIA, an interactive system that endows Claude Monet’s art with responsive auditory experiences. SOVIA uses computer vision trained on Monet’s artwork to take the user “into the painting.'' When the user interacts with a digital version of Monet's landscapes, their mouse positions are mapped to sounds that artistically represent the objects that the user is currently exploring in the art. These interactive musical journeys have the potential to make classical art more captivating for modern …

Embride: A Madhubani Art Creator, Jui Banik

Computer Science and Engineering Master's Theses

Creative machines are playing an increasingly important role in the art world and society at large. This gives rise to the need to consider the ethical dimensions of artistic machine agents. Whose art are we amplifying, or more importantly, who is being left behind? Computational Creativity offers the opportunity to bring awareness and visibility to art forms that have been systemically suppressed due to a range of longstanding societal biases. In this paper, we introduce EMBRIDE, a system that creates Madhubani art, practiced by women in the villages of India and Nepal. We hope that this work will encourage further …

Deep Learning Pseudocode Generation: A Qualitative Analysis, Colin Rioux

Computer Science and Engineering Master's Theses

Pseudocode is a traditional teaching tactic in computer science, yet it is not standardized and programming language dependent. Thus, it can be quite time consuming to write it. With the advancement of AI methodologies in NLP, AI could help address this problem. This work investigates the quality of AI generated pseudocode from source code. Five studies are conducted in this work to measure pseudocode quality, where each study modifies model input to observe accuracy and generalizability. The results show that there is an association between pseudocode quality and training and test set similarity. Furthermore, a sizable and diverse training set …

Learning With Your Buddies: An Investigation Of Community Based Ux Design Learning On Discord, Grace Ling

Computer Science and Engineering Master's Theses

Online communities have been a major part of how people connect with others to learn about different perspectives. In this thesis, I examine ways people use Discord, one of the major online community platforms, to learn UX design.

In this research, I designed a study, collected data from the Design Buddies Discord, and conducted semantic content analysis to investigate the community learners, job seekers, and mentors’ dialogues. I then used social network analysis to uncover patterns in connections. Lastly, I conducted a qualitative evaluation to survey and to understand the usefulness of Design Buddies.

The results show that a balance …

Applications For Nutrition Education In Developed And Developing Countries, Emma Allegrucci

Computer Science and Engineering Master's Theses

Food is vitally important for human beings. Without food, humanity would perish. Not only does food provide us with energy, but it also provides us with adequate nutrients so the systems throughout our body can function properly. Unfortunately, many people throughout the world, from top rated athletes to people living in impoverished areas, are either uninformed or do not have easy access to nutritional information or advice. There is a huge malnutrition epidemic among elite collegiate athletes and an even bigger malnutrition problem among the rural population of Uganda.

To solve the problem of malnourishment of collegiate athletes, I have …

Full Body Image Animation With Deep Learning: A Review, Rachael Brooks

Computer Science and Engineering Master's Theses

Deepfake technology has been undoubtedly growing at a rapid pace since 2017. Particularly since using GAN architecture was popularized, research in this area has grown and seems to only be gaining momentum. One interesting area is animating images of full body humans using deep learning. This paper looks at the research done in this area and research that can influence it by looking at papers regarding human pose transfer, human motion transfer, and human motion generation. All of these types of papers have similar requirements, where a target pose must be abstracted to a skeleton and combined with appearance data …

The Emily Dickinson Machine & Hybrid Poetry Generation, Juliana Shihadeh

Computer Science and Engineering Master's Theses

This thesis introduces EMILY, a machine that creates original poems in the style of renowned poet Emily Dickinson. Dickinson’s succinct and syntactically distinct style with unconventional punctuation makes for an interesting challenge for automated poetry creation. Furthermore, we adapt EMILY to answer the following hypothetical question: What if Emily Dickinson had collaborated with another poet from a different time period? To this end, we introduce Hybrid Generative Poetry, which simultaneously integrates poetic elements from multiple poets. Using two distinct approaches to Hybrid Poetry generation, we create poetry in the combined styles of Emily Dickinson and Robert Frost. User studies are …

Hydration Automation: Smart Tank, Peter Ferguson

Computer Science and Engineering Master's Theses

Agriculture is a huge field that can benefit from Internet of Things (IoT) with many solutions focused around the monitoring and management of water in irrigation or storage systems. Different types of sensors to measure water quality, amount, flow and pressure are used to monitor the health of these systems and actuators are deployed like pumps or valves to manage the water in theses systems. The Hydration Automation (HA) automated water monitoring and management system utilizes low cost, small form factor, and sustainable Sensing Units (SUs) to collect water level reading of water storage systems through the use of an …

Control Traffic And Queueing Latencies: Implicit Overheads In Bandwidth Sliced Software Defined Networks, Jesse Chen

Computer Science and Engineering Master's Theses

The southbound control protocols used in Software Defined Networks (SDNs) allow for centralized control and management of the data plane. However, these protocols introduce additional traffic and delay between network controllers and switches. Despite the well understood capabilities of SDNs, current representations of control traffic overhead consist of approximations at best. In addition to high reactivity to incoming flows, the need for resource allocation and deterministic messaging delay necessitates a thorough understanding and modeling of the amount of control traffic and its effect on latency. Software switching facilitates the development of edge and fog computing networks by allowing the use …

Deep Reinforcement Learning For Dialogue Systems With Dynamic User Goals, Glen Chandler

Computer Science and Engineering Master's Theses

Dialogue systems have recently become a widely used system across the world. Some of the functionality offered includes application user interfacing, social conversation, data interaction, and task completion. Most recently, dialogue systems have been developed to autonomously and intelligently interact with users to complete complex tasks in diverse operational spaces. This kind of dialogue system can interact with users to complete tasks such as making a phone call, ordering items online, searching the internet for a question, and more. These systems are typically created by training a machine learning model with example conversational data. One of the existing problems with …

Gaussian-Awareness Deep Learning For Block-Level Compressive Video Sensing, Yifei Pei

Computer Science and Engineering Master's Theses

Compressive sensing (CS) is a signal processing framework that effectively recovers a signal from a small number of samples. Traditional compressed sensing algorithms, such as basis pursuit (BP) and orthogonal matching pursuit (OMP) have several drawbacks, such as low reconstruction performance at small compressed sensing rates and high time complexity. Recently, researchers focus on deep learning to get compressive sensing matrix and reconstruction operations collectively. However, they failed to consider sparsity in their neural networks to compressive sensing recovery; thus, the reconstruction performances are still unsatisfied. In this thesis, we use 2D-discrete cosine transform and 2D-discrete wavelet transform to impose …

Deep Learning Methods For Efficient Image Coding, Zachary Daniel Bellay

Computer Science and Engineering Master's Theses

Video data makes up 58% of all internet traffic and is growing as self-driving car cameras, 4K televisions, and video surveillance systems continue to come online. Traditional heuristics based image and video codecs such as JPEG and HEVC have been successful thus far, however, these approaches lack the ability to leverage big data to gain massive insights. Six deep learning based approaches are proposed to tackle efficient image/video compression and image compression for machine classification.

Prioritized Anomaly Catalog Generation Using Model-Based Reasoning, Jake Hedlund

Computer Science and Engineering Master's Theses

Anomaly management—the detection, diagnosis, and resolution of anomalies in a system—is traditionally performed using experiential techniques which are quickly computed, but poorly structured. Newer model-based approaches are more systematic and higher performing but are computationally expensive, which is a particular challenge for execution in an operational environment. This paper builds on a novel system to pre-compute model-based anomaly symptoms to enable quick retrieval and diagnosis in operational settings. New additions to this system include a simplified model interface, anomaly likelihoods associated with each component, and easier interpretation of results. The implemented system has been used successfully to detect and diagnose …

Channel Scanning And Access Point Selection Mechanisms For 802.11 Handoff: A Survey, Dhananjay Singh

Computer Science and Engineering Master's Theses

While the cellular technology has been evolving continuously in recent years and client handoffs remain unnoticed, the 802.11 networks still impose an enormous latency issue once the client device decides to roam between the Access Point (AP). This latency is caused by many factors reckoning on scanning the channels and searching for APs with better signal strength. Once data from all the nearby APs has been collected, the client picks the most suitable AP and tries to connect with it. The AP verifies if it has enough capability to serve the client. It also ensures that the client has the …

Analysis Of The Duration And Energy Consumption Of Aes Algorithms On A Contiki-Based Iot Device, Brandon Tsao

Computer Science and Engineering Master's Theses

With the growing prevalence of the Internet of Things, securing the sheer abundance of devices is critical. The current IoT and security landscapes lack empirical metrics on encryption algorithm implementations that are optimized for constrained devices, such as encryption/decryption duration and energy consumption. In this paper, we achieve two things. First, we survey for optimized implementations of symmetric encryption algorithms. Seconds, we study the performance of various symmetric encryption algorithms on a Contiki-based IoT device. This paper provides encryption and decryption durations and energy consumption results on three implementations of AES: TinyAES, B-Con’s AES, and Contiki’s own built-in AES. In …

Extreme Image Compression With Deep Learning Autoencoder, Licheng Xiao

Computer Science and Engineering Master's Theses

Image compression can save billions of dollars in the industry by reducing the bits needed to store and transfer an image without significantly losing visual quality. Traditional image compression methods use transform, quantization, predictive coding and entropy coding to tackle the problem, represented by international standards like JPEG (joint photographic experts group), JPEG 2000, BPG (better portable graphics), and HEIC (high efficiency image file format). Recently, there are deep learning based image compression approaches that achieved similar or better performance compared with traditional methods, represented by autoencoder, GAN (generative adversarial networks) and super-resolution based approaches.

In this paper, we built …

The Fog Development Kit: A Platform For The Development And Management Of Fog Systems, Colton Powell

Computer Science and Engineering Master's Theses

With the rise of the Internet of Things (IoT), fog computing has emerged to help traditional cloud computing in meeting scalability demands. Fog computing makes it possible to fulfill real-time requirements of applications by bringing more processing, storage, and control power geographically closer to end-devices. How- ever, since fog computing is a relatively new field, there is no standard platform for research and development in a realistic environment, and this dramatically inhibits innovation and development of fog-based applications. In response to these challenges, we propose the Fog Development Kit (FDK). By providing high-level interfaces for allocating computing and networking resources, …

Overhead Management Strategies For Internet Of Things Devices, Kavin Kamaraj

Computer Science and Engineering Master's Theses

Overhead (time and energy) management is paramount for IoT edge devices considering their typically resource-constrained nature. In this thesis we present two contributions for lowering resource consumption of IoT devices. The first contribution is minimizing the overhead of the Transport Layer Security (TLS) authentication protocol in the context of IoT networks by selecting a lightweight cipher suite configuration. TLS is the de facto authentication protocol for secure communication in Internet of Things (IoT) applications. However, the processing and energy demands of this protocol are the two essential parameters that must be taken into account with respect to the resource-constraint nature …

Image Classification On Iot Edge Devices: Profiling And Modeling, Salma Abdel Magid

Computer Science and Engineering Master's Theses

With the advent of powerful, low-cost IoT systems, processing data closer to where the data originates, known as edge computing, has become an increasingly viable option. In addition to lowering the cost of networking infrastructures, edge computing reduces edge-cloud delay, which is essential for mission-critical applications. In this thesis, we show the feasibility and study the performance of image classification using IoT devices. Specifically, we explore the relationships between various factors of image classification algorithms that may affect energy consumption such as dataset size, image resolution, algorithm type, algorithm phase, and device hardware. Our experiments show a strong, positive linear …

Energy Measurement And Profiling Of Internet Of Things Devices, Immanuel Amirtharaj

Computer Science and Engineering Master's Theses

As technological improvements in hardware and software have grown in leaps and bounds, the presence of IoT devices has been increasing at a fast rate. Profiling and minimizing energy consumption on these devices remains to be an an essential step towards employing them in various application domains. Due to the large size and high cost of commercial energy measurement platforms, the research community has proposed alternative solutions that aim to be simple, accurate, and user friendly. However, these solutions are either costly, have a limited measurement range, or low accuracy. In addition, minimizing energy consumption in IoT devices is paramount …

Non-Mpm Mode Coding For Intra Prediction In Video Coding, Taru Kanchan

Computer Science and Engineering Master's Theses

The High Efficiency Video Coding standard introduced thirty-five intra prediction modes. It employed a method based on three most probable modes (MPM) to improve intra mode coding. This method significantly improved the performance by extracting three MPMs out of the thirty-five intra modes. The Joint Video Exploration Team (JVET) defines sixty-seven intra prediction modes for a possible future video coding standard. In the latest JVET development, six MPMs are chosen, and the remaining sixty-one modes are divided into sixteen “selected” and forty-five “non-selected” modes. These non-MPM modes are coded using fixed length coding. This research focusses on finding more efficient …

Panoramic Stereovision And Scene Reconstruction, Ashish Nair

Computer Science and Engineering Master's Theses

With advancement of research in robotics and computer vision, an increasingly high number of applications require the understanding of a scene in three dimensions. A variety of systems are deployed to do the same. This thesis explores a novel 3D imaging technique. This involves the use of catadioptric cameras in a stereoscopic arrangement. A secondary system aims to stabilize the system in the event that the cameras are misaligned during operation. The system provides a stark advantage due to it being a cost effective alternative to present day standard state-of-the-art systems that achieve the same goal of 3D imaging. The …

Code Girl, Amanda Holl

Computer Science and Engineering Master's Theses

Despite the growing importance of technology and computing, fewer than one percent of women in college today choose to major in computer science. Educational programs and games created to interest girls in computing, such as Girls Who Code and Made With Code, have been successful in engaging girls with interactive and creative learning environments, but they are too advanced for young girls to benefit from. To address the lack of educational, computer science games designed specifically for young girls, we developed a web-based application called Code Girl for girls age five to eight to customize their own avatars using Blockly, …

Extension Of Model-Based Fault Diagnosis Techniques For Network Systems, Ahmed Naail Abeer

Computer Science and Engineering Master's Theses

No abstract provided.

Integrated Home Server, Santiago John Rose

Computer Science and Engineering Master's Theses

Since the advent of the microprocessor in the 1970s, the market for consumer electronics has exploded with new devices changing the way we live and do business. Today, mobile phones, cameras, PCs, iPads, mp3 players, network media players, security systems, automation and IT systems, all have common functionality and there is an increasing need for unification of access to all these devices around a common server based architecture to unlock the benefits of smart integration and to simplify access for the end user.

IHS project is designed to provide to its business and home owners a unified network for all …

Toward A Fair Transactive Energy Market: A Deeplearning Based Energy Consumption Prediction Model, Yuka Hatori

Computer Science and Engineering Master's Theses

The application of machine learning is vast and quickly spreading across disciplines because of its versatile utility. By nature, machine learning implementations can quickly be obfuscated and ultimately introduce and perpetuate discriminatory practices, which leads to the issue of fairness. Transactive energy and the distribution of energy management technologies allow for new participants, meaning individual households and entities smaller than large energy providers, to enter the market to buy and sell energy. Machine learning has potential for meaningful use in many aspects of the transactive energy market process, and we focus on the specific aspect of how individual households can …