Computer Sciences Commons^™

Handwriting Transformers, Ankan Kumar Bhunia, Salman Khan, Hisham Cholakkal, Rao Muhammad Anwer, Fahad Shahbaz Khan, Mubarak A. Shah

Computer Vision Faculty Publications

We propose a novel transformer-based styled handwritten text image generation approach, HWT, that strives to learn both style-content entanglement as well as global and local writing style patterns. The proposed HWT captures the long and short range relationships within the style examples through a self-attention mechanism, thereby encoding both global and local style patterns. Further, the proposed transformer-based HWT comprises an encoder-decoder attention that enables style-content entanglement by gathering the style representation of each query character. To the best of our knowledge, we are the first to introduce a transformer-based generative network for styled handwritten text generation. Our proposed HWT …

Lecture 08: Partial Eigen Decomposition Of Large Symmetric Matrices Via Thick-Restart Lanczos With Explicit External Deflation And Its Communication-Avoiding Variant, Zhaojun Bai

Mathematical Sciences Spring Lecture Series

There are continual and compelling needs for computing many eigenpairs of very large Hermitian matrix in physical simulations and data analysis. Though the Lanczos method is effective for computing a few eigenvalues, it can be expensive for computing a large number of eigenvalues. To improve the performance of the Lanczos method, in this talk, we will present a combination of explicit external deflation (EED) with an s-step variant of thick-restart Lanczos (s-step TRLan). The s-step Lanczos method can achieve an order of s reduction in data movement while the EED enables to compute eigenpairs in batches along with a number …

Teaching Social-Emotional Skills Through Storytelling: Development Of A Mobile App For Children, Sally Devitry

Student Research Symposium

Children today are born into a world where technology is deeply integrated into daily life. A study completed in 2016 found that 85% of children ages 5-10 participate in some form of screen time (e.g., television, tablet, or smart phone) daily. More than 75% of these children do so for more than two hours a day on average. The increasing use of technology, specifically tablet and smartphone use, will fundamentally redefine childhood experiences. In a child’s formative years, learning social-emotional skills such as empathy, communication, resilience, etc. is vital. Obviously, human interaction cannot be replaced in teaching these social-emotional skills, …

Solving Classical Computing Problems Via Quantum Computing, David Wisnosky

Showcase of Osprey Advancements in Research and Scholarship (SOARS)

Honorable Mention Winner

The field of computing has in recent years begun to hit a wall in what is computationally feasible. In the past problems that were complex from a time standpoint waited for hardware to advance. Currently, Moore’s Law, the observation that the number of transistors on a chip doubles about every two years, has come to an end. This is due to the size of a transistor becoming so small that it begins to experience quantum effects and the laws of physics upon which a computer is built break down. A new paradigm of computing has emerged to …

Oer Immersive Mulitmedia Materials Project: Vr As An Agent Of Change, Alexander Isin, Jade Basilius, Johana Barrero, Angeles Fernandez-Cifuentes

Showcase of Osprey Advancements in Research and Scholarship (SOARS)

OER-Immersive Multimedia Materials Project: VR as an Agent of Change Alexander Isin, Computer Science/ Spanish Major (Main researcher) Jade Basilius, Communication/ Spanish Major (Main researcher) Mentors: Johana Barrero, Ph. D., Ångeles Fernández Cifuentes, Ph. D. This project aims at the design and development of immersive instructional materials (with a focus on virtual reality) for the Spanish language, literature, and culture classroom. It specifically involves the design of an interactive 360-degree experience in Spanish about the Florida Keys Coral Reefs and their environmental, socioeconomic, and cultural impact on the community. My involvement in the project focused on the research of different …

Garduino: Using Image Processing To Measure Health In Plants, Maria Pugliese, Matthew Tapia, Adam Flowers, Patrick Kreidl

Showcase of Osprey Advancements in Research and Scholarship (SOARS)

Honorable Mention Winner

As the world population increases, so do the demands for more efficient (less energy-consuming) methods of food cultivation. The vision of “precision agriculture” strives for greater farmland efficiency through advances in technology (sensors, robots), promising capabilities beyond what is possible from only manual labor. The University of North Florida’s “Garduino” project, providing a “hands-on” garden bed within which customized automated solutions can be piloted, aims to prepare engineering and computing undergraduates for this precision agriculture vision. A particularly valuable data source in this vision is near-field crop imagery, as may be acquired via self-navigating ground robots with …

Shor’S Algorithm: How Quantum Computing Affects Cybersecurity, Caroline Fedele, Asai Asaithambi

Showcase of Osprey Advancements in Research and Scholarship (SOARS)

Honorable Mention Winner

Almost all of today’s computer security relies on something known as the RSA cryptosystem. This system relies on a mathematical, specifically number theory, problem known as prime factorization, where a composite number is broken down into its two prime number factors. This in an ideal method for encryption because it is easy to multiply two numbers, encoding the data, but it much harder to determine which numbers were originally multiplied together, thus hard to decode the data. If this composite number is sufficiently large, there is no known algorithm for efficiently breaking it down – at least …

Distfold: Distance-Guided Protein Folding, Matthew Bernardini

Theses

Protein structure prediction and its associated key sub-problems such as distance map prediction are of significance importance in biology and bioinformatics. The inter-residue distance prediction problem, or distance prediction in short, is to predict the physical distance between amino acids in a three-dimensional (3D) space, given a protein's one-dimensional sequence information. While there exist many methods to predict distance maps, there are currently no methods that can take those predicted distance maps and build 3D models from them in an ab initio way, i.e., without using any other information. This works aims to fill this gap by: a) developing a …

Cyberbullying: Its Social And Psychological Harms Among Schoolers, Hyeyoung Lim, Hannarae Lee

International Journal of Cybersecurity Intelligence & Cybercrime

Criminal justice around the world has prioritized the prevention and protection of bullying and its victims due to the rapid increases in peer violence. Nevertheless, relatively few studies have examined what treatments or assistance are effective for peer victims to reduce and recover from their social and psychological suffering, especially in cyberbullying cases. Using data derived from the National Crime Victimization Survey-School Crime Supplement data in 2011 and 2013 (N=823), the current study examined the impact of two emotional support groups (i.e., adult and peer groups) on cyberbullying victims' social and psychological harm. The findings indicated that both adult and …

Cyber-Victimization Trends In Trinidad & Tobago: The Results Of An Empirical Research, Troy Smith, Nikolaos Stamatakis

International Journal of Cybersecurity Intelligence & Cybercrime

Cybertechnology has brought benefits to the Caribbean in the form of new regional economic and social growth. In the last years, Caribbean countries have also become attractive targets for cybercrime due to increased economic success and online presence with a low level of cyber resilience. This study examines the online-related activities that affect cybercrime victimization by using the Routine Activity Theory (RAT). The present study seeks to identify activities that contribute to different forms of cybercrime victimization and develop risk models for these crimes, particularly the understudied cyber-dependent crimes of Hacking and Malware. It also aims to explore if there …

Lecture 00: Opening Remarks: 46th Spring Lecture Series, Tulin Kaman

Mathematical Sciences Spring Lecture Series

Opening remarks for the 46th Annual Mathematical Sciences Spring Lecture Series at the University of Arkansas, Fayetteville.

Vlsi Implementation Of A Cost-Efficient Loeffler-Dct Algorithm With Recursive Cordic For Dct-Based Encoder, Rih-Lung Chung, Chen-Wei Chen, Chiung-An Chen, Patricia Angela R. Abu, Shih-Lun Chen

Department of Information Systems & Computer Science Faculty Publications

This paper presents a low-cost and high-quality; hardware-oriented; two-dimensional discrete cosine transform (2-D DCT) signal analyzer for image and video encoders. In order to reduce memory requirement and improve image quality; a novel Loeffler DCT based on a coordinate rotation digital computer (CORDIC) technique is proposed. In addition; the proposed algorithm is realized by a recursive CORDIC architecture instead of an unfolded CORDIC architecture with approximated scale factors. In the proposed design; a fully pipelined architecture is developed to efficiently increase operating frequency and throughput; and scale factors are implemented by using four hardware-sharing machines for complexity reduction. Thus; the …

Assessing The Credibility Of Cyber Adversaries, Jenny A. Wells, Dana S. Lafon, Margaret Gratian

International Journal of Cybersecurity Intelligence & Cybercrime

Online communications are ever increasing, and we are constantly faced with the challenge of whether online information is credible or not. Being able to assess the credibility of others was once the work solely of intelligence agencies. In the current times of disinformation and misinformation, understanding what we are reading and to who we are paying attention to is essential for us to make considered, informed, and accurate decisions, and it has become everyone’s business. This paper employs a literature review to examine the empirical evidence across online credibility, trust, deception, and fraud detection in an effort to consolidate this …

The Present And Future Of Artificial Intelligence In Ophthalmology, Robert Abishek, Elliot Cherkas

inSIGHT

Dr. Ravi Goel is a comprehensive ophthalmologist and cataract surgeon at Wills Eye Hospital, with a specific interest in finding ways that AI can help ophthalmologists improve their clinical care and treat more patients. Dr. Goel also publishes a daily blog, Protecting Sight, where he discusses a variety of topics ranging from advances in cataract surgery to medical education. One common thread throughout his blog is the burgeoning impact of AI on the field of ophthalmology, such as the utility of deep learning algorithms for diagnosing various diseases and the impact that improved intra-ocular lens (IOL) power calculations will have …

Iot Based Agriculture 4.0: Challenges And Opportunities, Halimjon Khujamatov, Temur Toshtemirov Mr., Doston Turayevich Khasanov Mr., Nasiba Saburova Ms., Ilhom Ikromovich Xamroyev Mr.

Bulletin of TUIT: Management and Communication Technologies

In recent years, the world's population growth has been intensifying, resulting in specific problems related to the depletion of natural resources, food shortages, declining fertile lands, and changing weather conditions. This paper has been discussed the use of IoT technology as a solution to such problems.

At the same time, the emergence of IoT technology has given rise to a new research direction in agriculture. Soil analysis and monitoring using Zigbee wireless sensor network technology, which is part of the IoT, will enable the creation of an IoT ecosystem as well as the development of smart agriculture. In addition, entrepreneurship, …

Why Decimal System? Why Communities With More Than 150 Folks Tend To Split? New Consequences Of The Seven Plus Minus Two Law, Leobardo Orea Amador, Vladik Kreinovich

Departmental Technical Reports (CS)

Why are we using the decimal system to describe numbers? Why all over the world, communities with more than 150 folks tend to split? In this paper, we show that both phenomena -- as well as some other phenomena -- can be explained if we take into account the seven plus minus two law, according to which a person can keep in immediate memory from 5 to 9 items.

Taiger Ai: Saas Bundling And Unbundling, Singapore Management University

Perspectives@SMU

Software companies bundle support services with their products as standard practice. Is it possible to be different…and profitable?

Predicting The Outcome Of Nba Games, Matthew Houde

Honors Projects in Data Science

The aim of the project is to create a machine learning model to predict NBA games. The purpose is to build upon and improve existing models. Research into other predictive sports models and machine learning techniques was conducted to understand what is currently being done to predict NBA games and how effective it is in doing so. After a thorough literary review, the model was created using Python and a variety of machine learning techniques. The dataset used had an array of team statistics for both the home and away team for each corresponding matchup and two supporting features were …

Data-Limited Domain Adaptation And Transfer Learning For Learning Latent Expression Labels Of Child Facial Expression Images, Megan Witherow, Winston Shields, Manar Samad, Khan Iftekharuddin

College of Engineering & Technology (Batten) Posters

While state-of-the-art deep learning models have demonstrated success in adult facial expression classification by leveraging large, labeled datasets, labeled data for child facial expression classification is limited. Due to differences in facial morphology and development in child and adult faces, deep learning models trained on adult data do not generalize well to child data. Recent deep domain adaptation approaches have improved the generalizability of models trained on a source domain to a target domain with few labeled samples. We propose that incorporating steps of deep transfer learning, e.g. weights initialization from the pre-trained source model and freezing model layers, may …

Analysis Of Reading Patterns Of Scientific Literature Using Eye-Tracking Measures, Gavindya Jayawardena, Sampath Jayarathna, Jian Wu

College of Sciences Posters

Scientific literature is crucial for researchers to inspire novel research ideas and find solutions to various problems. This study presents a reading task for novice researchers using eye-tracking measures. The study focused on the scan paths, fixation, and pupil dilation frequency of the participants. In this study, 3 participants were asked to read a pre-selected research paper while wearing an eye-tracking device (PupilLabs Core 200Hz). We specified sections of the research paper as areas of interest (title, abstract, motivation, methodology, conclusion)to analyze the eye-movements. Then we extracted eye-movements data from the recordings and processed them using an eye-movement processing pipeline. …

Nanopore Guided Regional Assembly, Eleni Adam, Desh Ranjan, Harold Riethman

College of Sciences Posters

The telomeres are the “caps” of the chromosomes and their vital role is to protect them. Possible telomere dysfunction caused by telomere rearrangements can be fatal for the cell and result in age-related diseases, including cancer. The telomeres and subtelomeres are regions that are hard to investigate. The current technology cannot provide their complete sequence, instead the DNA is given in multiple pieces. Current methods of assembling the pieces of these regions are not accurate enough due to the region’s high variability and complex repeated patterns. We propose a hybrid assembly method, the NPGREAT, which utilizes two of the latest …

Combine Cryo-Em Density Map And Residue Contact For Protein Structure Prediction: A Case Study, Maytha Alshammari, Jing He

College of Sciences Posters

Although atomic structures have been determined directly from cryo-EM density maps with high resolutions, current structure determination methods for medium resolution (5 to 10 Å) cryo-EM maps are limited by the availability of structure templates. Secondary structure traces are lines detected from a cryo-EM density map for α-helices and β-strands of a protein. A topology of secondary structures defines the mapping between a set of sequence segments in 1D and a set of traces of secondary structures in 3D. In order to enhance the accuracy in ranking secondary structure topologies, we propose a method that combines three sources of information …

Vaim For Solving Inverse Problems, Manal Almaeen, Yasir Alanazi, Michelle Kuchera, Nobuo Sato, Wally Melnitchouk, Yaohang Li

College of Sciences Posters

In this work, we propose the Variational Autoencoder Inverse Mapper (VAIM) to solve inverse problems, where there is a demand to accurately restore hidden parameters from indirect observations. VAIM is an autoencoder-based neural network architecture. The encoder and decoder networks approximate the forward and backward mapping, respectively, and a variational latent layer is incorporated into VAIM to learn the posterior parameter distributions with respect to the given observables. VAIM shows promising results on several artificial inverse problems. VAIM further demonstrates preliminary effectiveness in constructing the inverse function mapping quantum correlation functions to observables in a quantum chromodynamics analysis of nucleon …

The Power Of The "Internet Of Things" To Mislead And Manipulate Consumers: A Regulatory Challenge, Kate Tokeley

Notre Dame Journal on Emerging Technologies

The “Internet of Things” revolution is on its way, and with it comes an unprecedented risk of unregulated misleading marketing and a dramatic increase in the power of personalized manipulative marketing. IoT is a term that refers to a growing network of internet-connected physical “smart” objects accumulating in our homes and cities. These include “smart” versions of traditional objects such as refrigerators, thermostats, watches, toys, light bulbs, cars, and Alexa-style digital assistants. The corporations who develop IoT are able to utilize a far greater depth of data than is possible from merely tracking our web browsing in regular online environments. …

Sound In Color, Amber Rhodes

Honors Scholars Collaborative Projects

“Sound in Color” is an interactive audio-visual experience designed to explore the relationship between sound, color, and emotions. Taking place on the Massey Concert Hall stage, the project is inspired by synesthesia and incorporates research on color psychology. Participants are invited to select an emotion and color. As the user hums into a microphone, they hear their emotions expressed through sound in their headphones and watch as the lights on stage respond to their vocal cues.

An Education Theory Of Fault For Autonomous Systems, William D. Smart, Cindy M. Grimm, Woodrow Hartzog

Notre Dame Journal on Emerging Technologies

Automated systems like self-driving cars and “smart” thermostats are a challenge for fault-based legal regimes like negligence because they have the potential to behave in unpredictable ways. How can people who build and deploy complex automated systems be said to be at fault when they could not have reasonably anticipated the behavior (and thus risk) of their tools? Part of the problem is that the legal system has yet to settle on the language for identifying culpable behavior in the design and deployment for automated systems. In this article we offer an education theory of fault for autonomous systems—a new …

Multi-Dimensional Numerical Integration On Parallel Architectures, Ioannis Sakiotis, Marc Paterno, Balsa Terzic, Mohammad Zubair, Desh Ranjan

College of Sciences Posters

Multi-dimensional numerical integration is a challenging computational problem that is encountered in many scientific computing applications. Despite extensive research and the development of efficient techniques such as adaptive and Monte Carlo methods, many complex high-dimensional integrands can be too computationally intense even for state-of-the-art numerical libraries such as CUBA, QUADPACK, NAG, and MSL. However, adaptive integration has few dependencies and is very well suited for parallel architectures where processors can operate on different partitions of the integration-space. While existing parallel methods exist, most are simple extensions of their sequential versions. This results in moderate speedup and in many cases failure …

End-To-End Physics Event Generator, Yasir Alanazi, N. Sato, Tianbo Liu, W. Melnitchouk, Michelle P. Kuchera, Evan Pritchard, Michael Robertson, Ryan Strauss, Luisa Velasco, Yaohang Li

College of Sciences Posters

We apply generative adversarial network (GAN) technology to build an event generator that simulates particle production in electron-proton scattering that is free of theoretical assumptions about underlying particle dynamics. The difficulty of efficiently training a GAN event simulator lies in learning the complicated pat- terns of the distributions of the particles physical properties. We develop a GAN that selects a set of transformed features from particle momenta that can be generated easily by the generator, and uses these to produce a set of augmented features that improve the sensitivity of the discriminator. The new Feature-Augmented and Transformed GAN (FAT-GAN) is …

Why, In Deep Learning, Non-Smooth Activation Function Works Better Than Smooth Ones, Daniel Cruz, Richard Godoy, Vladik Kreinovich

Departmental Technical Reports (CS)

Since in the physical world, most dependencies are smooth (differentiable), traditionally, smooth functions were used to approximate these dependencies. In particular, neural networks used smooth activation functions such as the sigmoid function. However, the successes of deep learning showed that in many cases, non-smooth activation functions like max(0,z) work much better. In this paper, we explain why in many cases, non-smooth approximating functions often work better -- even when the approximated dependence is smooth.

Why Semi-Supervised Learning Makes Sense: A Pedagogical Note, Olga Kosheleva, Vladik Kreinovich

Departmental Technical Reports (CS)

The main idea behind semi-supervised learning is that when we do not enough human-generated labels, we train a machine learning system based on what we have, and we add the resulting labels (called pseudo-labels) to the training sample. Interesting, this idea works well, but why is somewhat a mystery: we did not add any new information so why is this working? There exist explanations for this empirical phenomenon, but most these explanations are based on complicated math. In this paper, we provide a simple intuitive explanation.