Zoom links will be active from 9:15 AM - 11:15 AM the day of the expo.
Jessica Hotsko, Chris Louviere
Zoom link: inactive
Project links: http://www.afrotsee.com/
https://github.com/jhotsko/AFROTSee.git
https://drive.google.com/drive/folders/1fkggzqdCBajT9mW-uhb-aecjqic8PQhT?usp=sharing
A website that helps juniors and seniors manage information/data for sophomores and freshmen within Air Force ROTC.
Mentor: Dr. Anastasia Kurdia
Christina Campbell, Mason Caplin, Sam Eigen, Jake Kalinsky, Cappy Lassen, Oliver Martin
Zoom link: inactive
Project link: https://app.box.com/s/z0r4j07f9wzstntb7glyiilv35n9uw61
For our project, we are taking a new and revolutionary approach to tackling the global crisis that is climate change. With this method, we will be able to appeal directly to the individual, giving them a personalized experience to delineate how and why climate change is so important. In a virtual reality world, an individual will traverse different scenes that represent different geographic regions. We decided to focus on one region Tulane students are most familiar with, the Southeast. Different aspects of climate change will slowly appear as the user travels through these environments. The climate change data is extrapolated through NASA and different government sites and all of the virtual reality is coded within Unity. The goal is simple: to prove to the user that no matter their political view, climate change is a crucial problem that must be tackled with data and science. By adding the virtual reality element, this experience becomes highly personal -- as the individual enters their own, seemingly failing world. The end of the VR space will represent a voting booth, showing statistically the best local, state, and federal officials to elect to help combat climate change, which will end on an inspiring note.
Sponsor: Schlumberger
Mentor: Dr. Aaron Maus
Alex Buck, Thomas Huang, Arjun Sree Manoj
Zoom link: inactive
Project links: https://github.com/tthuang1998/Capstone
https://app.box.com/s/j4a69gyvo5dfyhg2j0s0kjvap2zqitdj
Street art & graffiti are important forms of expression widely used in the New Orleans area. Due to the volatile nature of street art, it is hardly immortalized. This semester, street artist B-Mike has collaborated with the Newcomb Art Museum to put on an exhibit called NOT SUPPOSED 2-BE HERE. Our goal with our capstone project is to immortalize his work by preserving his works of art from the exhibit in a virtual space in a high-resolution and traversable 3D model using consumer-grade technology.
Sponsors: Tulane Computer Science Department, Schlumberger Design Project Fund
Mentors: Dr. Ramgopal Mettu, Brandon "B Mike" Odums
Kevin Au, Michael Montgomery, Brianna Tucker
Zoom link: inactive
Project link: https://docs.google.com/presentation/d/1tErXyEUY_C0OSAU5Mdd4eWWSIOQN96HWOQRILVwaqiA/edit?usp=sharing
Current competitive matchmaking ranking systems only consider whether a player’s team won or lost when granting player's rank, which may unfairly punish those with poorly preforming teammates or unjustly reward those who do not participate as much as their teammates.
Our main goal is to create a ranking system that counters theses issues by factoring in the individual contributions of each player in a team by assigning a “performance score” to weigh how important each member's work was to achieve the final result.
Mentor: Dr. Michael Mislove
Vinny Sgarzi, Alex Todorovic, Alexa Westlake
Zoom link: inactive
Project link: https://cointoss.live
Coin Toss is a long short-term memory, deep learning model capable of forecasting exchange rates for currency pairs and generating short term, multi-day trading strategies.
Mentor: Dr. Zizhan Zheng
Chanho Lim, Roland Nguyen
Zoom link: inactive
Project link: https://rjn-2b.github.io/DMG-Capstone/
Deep neural network GAN generating classical music in the style of a composer and classifier model to classify if the music can be determined to have come from a composer.
Mentor: Dr. Jihun Hamm
Jake Gartenstein, Ted Hoeller, Addie Jasica, Andrei Mills, Mason O'Connor
Zoom link: inactive
Project link: https://drive.google.com/drive/folders/16XIBSM5OgVVbZmwtLwBMHW8gzju8FEmq?usp=sharing
Bringing together the disciplines of computer science and dance, our project is focused on dance sequence generation for the human body based off any provided song from the Spotify streaming platform. It utilizes machine learning via neural networks, hours of dance footage, and the spotify API to bring in music data. We aim to inspire new and exciting choreography and at the same time produce a visualization that is in and of itself a piece of art.
Sponsor: Newcomb Tulane College - Timothy Sykes Daytrading Award for the Talented
Mentors: Dr. Richard Snow, Dr. Brian Summa
Hanyu Lu
Zoom link: inactive
Project link: https://drive.google.com/open?id=1Jd9H9uOidyDZ6DdI_HExaAW1RiLygIin
Introducing model poisoning attack in federated learning and a defense strategy to minimize its effectiveness.
Sponsor: Tulane University
Mentor: Dr. Zizhan Zheng
Margaux Armfield, Kennedy Dorsey
Zoom link: inactive
Project link: https://tulane.box.com/s/6iw3ekgvsx474ptvsq07jf27sk03k8d5
Our capstone project focused on enhancing FishNet2, a biodiversity portal that contains data on fish specimens from collections around the world. With the help of our advisors, Dr. Kurdia, Dr. Bart, and Dr. Bakis, we aimed to aid ichthyology researchers by improving the accessibility and usability of Fishnet2’s website, creating new data analysis tools, and increasing the accuracy of Fishnet’s data. We achieved these goals by gathering user-feedback from current Fishnet2 users, updating the website to be more intuitive and user friendly, and creating an R package specifically tailored to process and analyze the data obtained from FishNet2. Our hope is that these modifications and new features will support researchers in their mission to advance scientific knowledge about fish diversity.
Mentors: Dr. Yasin Bakis, Dr. Henry Bart, Dr. Anastasia Kurdia
Rosalind Kidwell, Ilana Seidl, Chase Stockwell
Zoom link: inactive
Project link: https://app.box.com/s/kuita0qs379j9mhfih5z5j0i9qpcsi3y
Our project explores the language and dialogues surrounding Cryptocurrency. Traditionally, people who want to learn about Cryptocurrency use online Reddit forums. Our group explored how these forms use gendered language skewed towards men, thus making it more difficult for people of other genders to learn more about Cryptocurrency.
Mentor: Dr. Nicholas Mattei
Shane Westerfer
Zoom link: inactive
Project link: https://app.box.com/s/0g29eu811soub89duqppz23gtntbuxzt
Runs a gymnastics routine against the requirements in the official code of points. Determines whether or not routine satisfies all requirements and if not, which requirements are unsatisfied.
Mentor: Dr. Nicholas Mattei
Ben Gertz, Jordan Kolbert
Zoom link: inactive
Project link: https://app.box.com/s/h5sx83yi1o0i723ah1w02lr5egjh5jko
Done in collaboration with Rmark bio, our project focuses on identifying levels of influence various doctors have within the medical community. Frequently, companies and committees seek members of the medical community who have influence in order to discuss agendas, or the benefits of specific medications. We developed a program using data-points such as grant applications, Medical Trials, Speaker Conferences, and more in order to allow users to identify which doctors will have had the most influence or interaction with a given group of doctors. This essentially operates as a recommender system, returning users a list of doctors that are influential amongst those within their given list. Our algorithm allows custom weighting of various data points, allowing users to emphasize certain factors that they consider more important for their personal use of the program.
Sponsor: Rmark bio
Mentor: Dr. Anastasia Kurdia
Joel Hochman, Olivia Nye, Tim Smith
Zoom link: inactive
Poster link: https://gofile.io/?c=oeRsHD
Mind is a mental health application that lets you log mood data and securely share that data with your mental health provider. For the capstone, we developed the Therapist Dashboard, which is the software that mental health providers use to view and manage their patients' data.
Sponsor: Schlumberger
Mentor: Dr. Ramgopal Mettu
Harrison Pratt, Thomas Roginsky
Zoom link: inactive
Project link: https://app.box.com/s/7v8s00v9i5m3c31c8to2q8zh2fcs33bm
Moving Target defense is a cybersecurity strategy that allows computers on a network to hide from an attacker by constantly changing their state so that they cannot be targeted. This project consists of an improvement to an existing algorithm that models this strategy as well as a demo that serves as a test bed for the algorithm.
Sponsor: National Science Foundation
Mentor: Dr. Zizhan Zheng
Benjamin Smallson Jackson
Zoom link: inactive
Project link: https://app.box.com/s/duj0fmqvhztq8fhdm0688qi2sb2ok5zw
Nano-Satellite Scheduling for NASA's EM-1 mission
Mentor: Dr. K. Brent Venable
Joseph Allen, Samuel Beebe
Zoom link: inactive
Project link: https://docs.google.com/presentation/d/1WMzukH5qTzAEkbGowgi3pZY6e3tKWRTTCJoMFngpSUc/edit?usp=sharing
An interactive, 3D visualization of the 100-year flood event in New Orleans. The visualization combines point cloud data from LSU's Atlas LiDAR (for the terrain) with shapefiles from FEMA containing the geometry of the floodplain (for the water).
Sponsor: Tulane University
Mentor: Dr. Brian Summa
Zach Seymour, Zekun Wu, Sarah Xing
Zoom link: inactive
Project link: https://app.box.com/s/d2mpah79t4davnwheufngl7eswulgg59
We made a website that pulls up articles in politics based on the user's search. It then provides a short summary of the article and uses sentiment analysis to determine if the article is left-leaning, right-leaning, or center.
Mentor: Dr. Zizhan Zheng
Jack Green, Dylan Hecht, Ben Johnson, Blake Zaffiro
Zoom link: inactive
Project link: https://app.box.com/s/b3ywe9tx9trvp47rojpjmr0gxtugjvl2
Our project is an investigation to determine the efficiency gains when replacing a gantry-based extruder in 3D printing with a robot arm. A robot arm allows additional degrees of freedom for the extruder and is paired with the use of a local search algorithm to determine if there is a significant difference in non-extrusion distance, print time, path computation time, swept volume of the arm, and energy expenditure. We carried our research out virtually through the use of a motion planning platform V-REP, with a robot arm with seven degrees of freedom.
Mentor: Dr. Ramgopal Mettu
Conrad Leonik
Zoom link: inactive
Project link: https://git.tulane.edu/cleonik/skourascore
In the brain, regions that are more activated pull more oxygen from the blood. Blood with different amounts of oxygen in it behaves differently under a magnetic field, and an fMRI machine utilizes this to create images of the brain that indicate which brain regions are activated at specific points in time. fMRI Neurofeedback is a task done by the test subject while they are in the MRI scanner, in which they are shown how activated a specific region of their brain is in real time, and they are instructed to attempt to control it, either by upregulating or downregulating. Research has shown that with practice, most people are able to significantly improve at this task over time. They are able to get better at controlling a specific region of their brain on command.
The goal of my project is to be able to predict whether a given person will be able to learn and improve at this task. I have developed an algorithm that can incorporate several sources of data, such as demographic information, resting-state fMRI scans, and various physiological tests. So far, through a classification approach, the algorithm is able to predict with about ~83% accuracy (AUC score = ~0.768) whether a given subject (who does not present with psychiatric pathology) will be able to meaningfully perform the Neurofeedback Task.
Mentors: Jeremy Cohen, Jihun Hamm, Jeffrey Rouse
Thomas Girmay
Zoom link: inactive
A convolutional neural network (CNN) that takes in images of individual items of garbage and returns its category: Paper, Plastic, Metal, Cardboard, Glass, or Trash.
Mentor: Dr. Nicholas Mattei
