In today's world, laying healthy foundations for a successful marriage is more crucial than ever in romantic relationships. Here, the "Before I Do" mobile application comes into play, focusing on couples in the pre-marital phase and guiding those who seek to strengthen their relationships through a scientific and objective approach.
Module 1: Relationship Assessment
This module provides couples with an in-depth inventory to understand themselves and their relationships. Covering four fundamental areas—personality traits, family dynamics, relationship dynamics, and cultural influences—couples can conduct an objective evaluation.
Module 2: Relationship Report
Couples who complete the inventory receive a relationship report presenting a scientific assessment of results. This report helps couples understand areas for improvement, guiding them in laying the groundwork for a healthy marriage.
Module 3: Expert Assistance and Self-Help
The third module offers couples the opportunity to access needed expert assistance and explore self-help resources. If desired, this module directs them to professionals and supports them with specialized content, creating a secure counseling environment. Based on inventory results, this module provides tailored content to help couples enhance areas where they show low to moderate compatibility. This empowers couples to become conscious builders of a healthier foundation for their relationship.
"Before I Do" offers a unique mobile experience, providing couples with the necessary tools for a conscious marriage. Are you ready to strengthen your relationship? Take a step towards a healthier future with "Before I Do."
Lola's 2.0 will be a grocery platform that uses artificial intelligence to create a customized food environment that promotes healthy, sustainable foods and drinks tailored to customer's budget and taste preferences.
Research to design and evaluate food policy relies on large datasets of food products which include many attributes (price, nutrients, ingredients, marketing elements, environmental sustainability); historically, processing, categorizing, and linking this data has been enormously time consuming and burdensome, as well as ultimately limited by the small number of attributes that can be considered at one time. This narrow focus on a single variable at a time has limited our ability to fully evaluate the impacts of implemented policies as well as design maximally effective policies, since being able to target the attributes and their combinations that consumers care about most would help make nudges, interventions, and policies more effective (e.g. incorporating price or front-of-package marketing elements into our models for how consumers make decisions in response to labels).
To address these gaps, we propose to develop an artificial intelligence (AI) model for food and nutrition data that will allow us to 1) expedite the processing time to review and categorize products for policy evaluation; and 2) create an AI-powered, personalized version of our experimental online food store (Lola’s 2.0) that will allow us to test and develop policies that are responsive to a variety of factors (including but not limited to nutrition, sustainability, and price), depending on a particular country’s policy and advocacy goals. We foresee potential applications in current focus countries and future additional countries. Together, this new AI approach for collecting, organizing, and categorizing food and nutrition data and testing policy design via Lola’s 2.0 will allow us to have a greater policy impact.
Client currently is doing research with a commercial system "Lola's 1.0". This proposal is to develope a new system from scratch and also making it customizable. The first step would be creating the data infrastructure for all the food and beverages products to categorize them by nutrition, price, brand and other attributes (Aim 1 in the proposal attached) so that those attributes can be used to create a customizable food shopping environment (Aim 2).
Client has Amazon Web Service credits that could be used to support the project. Also, I recognize the idea is ambitious- it would be fabulous if we only did part of it (for example, creating an app that promotes healthy, sustainable selections within a single food category; forgoing the "customizable" aspect of it, etc). I'm flexible! Thanks for the opportunity to submit this idea.
full project descriptionThe goal of TeachGenie is to utilize AI to solve the various problems teachers face. For now it aims to reduce the total time educators take to design and create teaching materials. TeachGenie generates lesson plans, activities, tests, and kahoots tailored to their classroom and students. It allows users to specify what topics, skills, standards, and state to focus the materials to best fit their needs.
The website is a full stack responsive application but there are plans of building a mobile app in the future. The technologies used include Angular(HTML, CSS, JS, TS), sqlite, fastapi, and Python. Students can speak to clients or otherwise come up with ideas on improving the application, and they can continue to build out the site.
more informationAs I mentioned by email, I'm happy to supervise a team to do extensions to the grad tracking software. One easy place to start is more automation and consistency in approving background requirements. The system is used to help grad students and department admin to keep track of the many milestones a student must accomplish before completing degree studies.
The team will be working with Dr. Porter, as well as users of the system (like Denise Kenney, and Dr. Jasleen Kaur).
I would like to work with a team to develop an app to automate the task of course scheduling for the Dept of Computer Science.
Background: Each semester, I have to perform course scheduling for the department, which involves around 60 courses, 40 instructors, 20 meeting patterns, and 30 classrooms. There are lots of constraints and preferences to accommodate as well. I would like a software tool to automate this task, even if it requires some small final manual cleanup.
Task Detail: This task entails (i) choosing which courses are to be offered; (ii) who will teach what; (iii) accommodate instructor's time preferences; (iv) determine a good schedule that minimizes conflicts and crowding; and (v) find classrooms appropriate to the enrollment size.
Algorithm Formulation: I have a general idea that the core of this problem might lend itself to an integer linear programming (ILP) formulation. In particular, there are constraints to be satisfied (this course must be offered, this course is optional, this instructor cannot teach in the afternoon, etc.), and there are things to be optimized (minimize overlaps/crowding, meet ideal class sizes, room not too big nor too small, etc.). So, this is a constrained optimization problem. We should be able to leverage open-source ILP solvers; in that case the problem becomes transforming the data into an ILP problem.
User Interface: It would be nice to have a simple interface to specify instructor's preferences and other constraints. At a bare minimum, a text file or spreadsheet will suffice. If there is time and interest, we could use a nicer data entry interface.
Platform: The tool does not need to be mobile. A web based app might be nice so it can be used from different computers. Something that could pull data from a spreadsheet (Google Sheet) would be nice though not required. A Google sheet has the benefit of being attached to a form where preferences and constraints and lists etc. can be populated.
I would like to work with a team to develop a tool to automate the semesterly task of matching students with jobs in the Dept of Computer Science.
Background: Each semester, there are 4-5 of us who have to perform the complex task of filling RA/TA/LA positions for all of our research projects and courses. This task is a complex matching problem, esp. before the start of the fall semester, when it can involve around 50-60 graduate students, 20-30 research projects, 20-30 graduate TA positions, and around 100 LA positions. There are lots of constraints and preferences to accommodate as well. I would like a software tool to automate this task, even if it requires some small final manual cleanup.
Task Detail: This task entails (i) first matching graduate students with RA positions, as much as possible; (ii) for those grad students who did not find a good match with an RA position, placing them as TAs in appropriate courses; (iii) hiring as many LAs as required to fully staff the courses. There are constraints (e.g., a TA/LA should have taken the course before) and preferences to accommodate.
Algorithm Formulation: I have a general idea that the core of this problem might lend itself to an integer linear programming (ILP) formulation. In particular, there are constraints to be satisfied (how many TA/LAs for a given course size, all graduate students with a funding guarantee must be matched with a job), and there are things to be optimized (instructors' and students' preferences). So, this is a constrained optimization problem. We should be able to leverage open-source ILP solvers; in that case the problem becomes transforming the data into an ILP problem.
User Interface: It would be nice to have a simple interface to specify instructor's and students' preferences and other constraints. At a bare minimum, a text file or spreadsheet will suffice. If there is time and interest, we could use a nicer data entry interface. We may also be able to integrate with a system developed in the XL lab for user input.
Platform: The tool does not need to be mobile. A web based app might be nice so it can be used from different computers. Something that could pull data from a spreadsheet (Google Sheet) would be nice though not required. A Google sheet has the benefit of being attached to a form where preferences and constraints and lists etc. can be populated.
We seek to expand a website developed by a COMP 523 team (S24). The website is running a Unity viewer which displays the 3D model of the Swayambhu temple environment that we created in 2022. We have a new version of this model and plans to continue to refine it. We need a fully functional website that incorporates a landing page, a model viewer that will allow users to explore an annotated model in a way that is true to life and intuitive to navigate (i.e., trackpad/mouse), additional contextualizing sub- pages, and the possibility of linking to the video database. This will entail work within Unity.
This project involves documenting and preserving Buddhist cultural heritage in Nepal, using annotated VR models. Your coach this semester (Sam Shi) was a member of the original project team. Thus you have good help for this proposed continuation.
Last semester, students worked to create a video database that uses a LLM to transcribe and translate video interviews from Nepal (in several languages). We want to expand the database and improve user functionality. The most important feature of this database is being able to search for text within the transcription. However, it must be optimized to be public-facing. Hence, we're also interested in building an intuitive user-interface.
For an accessible description of the project overall, see:
project coverage
Science Olympiad is a national organization that holds science tournaments for middle and high school students. This page has a brief description together with a couple of short videos to give you a better flavor for the program:
https://www.virginiaso.org/what-is-so
Events span the gamut of science and engineering fields, including biology, ecology, earth science, chemistry, physics, and mechanical and electrical engineering. However, computer science and software engineering events are not as well represented. One reason is the lack of a suitable computer-based environment in which students can solve coding problems under tournament conditions. We envision something like Gradescope, but addressing the unique challenges presented by Science Olympiad, including:
• It should be free, since SO tournaments have very limited funds.
• It should enable fast grading, because tournaments typically present awards the same
day that the tournament is held.
• It needs to allow students to compete in teams of two or three, with each member of
the team on their own computer, but able to see their own and their teammates’ answers
in the test.
• It needs to protect information about students, who are usually minors.
• It needs to suppress the temptation to cheat where possible. Cheating in this context
includes finding solutions on the internet as well as accessing or changing the answers
of a different team.
We envision a web-based testing environment, where the event judge (called an Event Supervisor, or ES) sets up a Wireless Access Point (WAP) connected to a computer that is running the testing environment’s server. Neither WAP nor computer should be connected to the Internet. Students will connect their computer to the WAP’s WiFi network and access the testing environment through their web browser. (The use of the WAP is the primary method for discouraging cheating, because this makes it difficult for students to connect to the open internet.)
Desirable features for the testing environment include:
• Relatively easy loading of questions with a minimal level of formatting.
. Auto-grading of answers by executing those answers and comparing the actual output to
the desired output. It should also capture any syntax or runtime errors, and it should
make answers and all outputs accessible to the ES.
• Support for problem solutions written in Python. Ideally, other languages should be supported
as well, but again this could be left until a subsequent version of the testing software.
• Provision for reference materials in web-compatible formats (HTML, PDF) to be loaded into,
and served by, the testing environment in addition to the test itself.
• Relatively easy creation and distribution of account logins to teams
Should this project be successful, we (Virginia Science Olympiad, or VASO) will make it freely available to all tournaments in Virginia, and once we deem it stable and mature enough, to all SO tournaments nationwide.
One of the most recent computer science/software engineering events run in the state of Virginia is an event called Cybersecurity, and the ES was Professor Saba Eskandarian in the UNC CS department. He is the person who recognized the need for this environment, and he will be collaborating with us on the requirements. He will also be using the MVP to run the Cybersecurity event at VASO tournaments this season to test it out and understand how we might improve it in the future.
Buckner Heavylift Cranes is a crane company based out of Graham, NC. In the fall of 2023, we worked with a COMP 523 team to create Buckner University, a website made to address our need to manage job training for our employees. Buckner uses a training tool called ITI that provides online courses on safety for our operators and employees.
We are looking for a new software system now. We are looking for a team of students to build an online reservation system for our conference rooms. The systems should allow any office worker to schedule time to use a conference room and show a calendar schedule of when the room is being used. Each conference room has an iPad outside to show the schedule so separate calendar pagers for each room will need to be displayed on the ipad. Additionally, the systems should send email confirmations of registration and a reminder of the scheduled time.
Key features: -- Schedule to display conference room availability and reservations -- Reservation system -- Integration with microsoft SSO login so only Buckner employees can access the system
Tech requirements: -- Node.js -- React frontend -- Express backend -- DB- MSSQL server -- Containerized with Docker -- Deployed to AWS app runner
Note: This is the tech stack of our other internal web apps and we would be open to utilizing different technology if the teams has a strong preference
Optional feature: Develop an integration with Outlook to link the conference room schedule with employees personal schedules
SympTalk is an AI-driven health assistant designed to bridge language barriers in healthcare by offering multilingual symptom assessment and care recommendations. This mobile app, with a web interface, aims to improve accessibility for underserved populations, particularly non-native speakers such as immigrants and refugees in the U.S. The app uses natural language processing (NLP) to support both text and voice inputs in multiple languages, translating symptoms into medically relevant terms for accurate analysis. AI-driven algorithms then provide tailored recommendations ranging from self-care advice to guidance on seeking medical attention. By integrating translation APIs like Google Translate or Microsoft Azure Translator, SympTalk ensures broad language support. The project addresses a critical gap in current healthcare tools, which often exclude common non-English languages, leading to misdiagnosis and delayed treatments. By enabling users to make informed decisions about their health, SympTalk democratizes healthcare, improving communication between patients and providers, and fostering better health outcomes in multilingual contexts.
Writers of narratives (e.g., fictional/nonfictional stories and books) often struggle to edit their work. In particular, they try to "show, not tell" in a piece of writing. What is the difference between "showing" and "telling?" Creative writing teachers often describe "showing" as putting the reader into the scene and "telling" as summarizing the scene. For example, a character's emotions could be described as happy/sad/angry/joyful/etc. (telling), or the character's emotions could be inferred with a description of his/her/their facial expressions (showing). Many times, the distinction between these qualities is nuanced and requires expertise and experience.
This semester I am researching ways in which LLMs can be used to measure the amount of showing vs. telling in a piece of text. As a key component of the final product, I would like to make publicly available a software that uses the results of the LLMs analysis to visually highlight areas in a text that are predominantly "showing" and/or "telling." This software should include a user interface in which the user inputs the text, and receives the same text back with colored highlighting and explanations for the highlighting choice.
I envision a computer application (or potentially a web application) which connects to an LLM (e.g., ChatGPT) via a personalized access token. Potential users include both teachers and writers. For example, instructors in the UNC Department of English have indicated interest in such a system if it is available.
The Dermatology Society Outreach Website is envisioned as a platform that raises awareness of dermatological health issues, with a particular emphasis on melanoma prevention and detection. This website will serve as a central hub for education, resources, and community engagement, targeting diverse audiences, including patients and advocates worldwide. The website's primary goal is to promote skin health awareness and prevention strategies, especially in populations at higher risk of melanoma or those often overlooked in traditional healthcare systems. By integrating multimedia resources, interactive tools, and community-driven initiatives, the platform will provide accessible, impactful content that inspires action and fosters collaboration across geographic and cultural boundaries. Application Domain This project will be a web-based application accessible via desktop and mobile browsers. While the primary access point is the website, it will be optimized for mobile devices to ensure reachability for users in all settings, particularly in low-resource areas. Additionally, the platform will offer integration with social media and external dermatology tools to maximize its impact. Target Audience The website will serve: -Patients and Families: Offering resources on melanoma prevention, early detection, and treatment options -Healthcare Professionals: Providing up-to-date research findings, case studies, and training materials -Community Advocates: Equipping leaders with tools to organize sun safety campaigns and educate their communities -Students: Sharing findings from The Dermatology Society’s initiatives Core Problems Addressed --Lack of Awareness: Educate users about melanoma and skin health in diverse skin tones which is a critical area often underrepresented in traditional campaigns --Global Accessibility: Provide free and multilingual resources for communities in underrepresented regions --Community Engagement: Empower advocates and healthcare workers to lead local awareness campaigns using the resources provided Website Pages and Features 1. Home Page Purpose: Introduce the mission and impact of The Dermatology Society Features: -Engaging banner with melanoma awareness statistics and calls to action -Quick links to key sections (ex. "Learn About Melanoma," "Get Involved," "Find Resources"). -Testimonial carousel from patients and healthcare professionals 2. About Us Purpose: Share The Dermatology Society’s mission, vision, and team members. Features: -History of the organization -Profiles of board members and volunteers -Annual reports and milestones achieved 3. Learn About Melanoma Purpose: Educate visitors about melanoma risks, prevention, and treatment Features: Interactive graphics showing melanoma development in different skin tones (multilingual) FAQs on common misconceptions about skin cancer Links to trusted medical journals and articles 4. Resources for Patients Purpose: Provide tools and guidance for individuals concerned about skin health Features: Skin self-exam tutorials with step-by-step videos Downloadable sunscreen use charts and UV index guides 6. Research Purpose: Highlight ongoing research and advocacy campaigns. Features: Impact stories from global initiatives 7. Community Engagement Purpose: Encourage visitors to get involved in the cause. Features: Volunteer form Campaign kit downloads, including flyers and templates Pictures of impact 8. Blog and News Purpose: Share picture, stories, updates, and expert opinions. Features: Articles written by dermatologists and advocates Updates on melanoma-related breakthroughs 9. Contact Us Purpose: Provide an avenue for communication and collaboration. Features: -Contact form for general inquiries -Interactive map showing global outreach -Telemedicine Tools: Link patients to teledermatology services for virtual consultations Data Analytics: Use analytics to track user engagement to optimize the site’s content
Cheeky Shopping is a social shopping platform offering a personalized, fashion-oriented shopping experience to our end users. Our web application is in the process of being released for a beta, and our mobile application is still in development. The application represents a blend of social media and e-commerce sites, providing a place where you can save or buy products you like and share those products with your friends.
Within the application, the Wardrobe acts as a place that a user can group products together and assign them as outfits. Unfortunately, product images from different brands are not easily displayed next to each other as a cohesive outfit. With the rise of generative AI and computer vision models in the technology industry, there is a unique opportunity to leverage these tools to solve this issue. As one of our features, we would like the user to be able to upload an image of themselves and one or more selected product images and receive an image of the user wearing the selected products.
In terms of architecture, our software is built cloud-native, with our current infrastructure on AWS. In order to fit into that ecosystem, the service should be built as a containerized microservice (using tools such as Docker) that can be easily deployed to cloud services.
Below is an example of a paper that describes the process that we are looking for called TryOnDiffusion. Diffusion models such as this require huge amounts of processing power and training data to build from the ground up, and we do not have the budget or resources to provide you with that. Instead the main goal for the project is to take this model or a model similar to it, and fine-tune the model to give more accurate and consistent results.
Link to the TryOnDiffusion paper: https://tryondiffusion.github.io/For my main job, I work for Microsoft as a Software Engineer.
We need a team to develop a software system that will employ several different machine learning (deep learning) techniques to assist in the recognition of different forms of lesions in dental cone beam computed tomography (CBCT) volumes (radiological images and 3D models).
The development of a method for automatic lesion detection and segmentation (determination of the boundaries of a lesion) may eventually improve dental practitioners’ ability to identify potentially clinically significant findings in cone beam computed tomography volumes requiring further follow-up or referral for management.
The dental medical details of the application will be handled by the client. Over 200 images are available that have been manually classified and can be used to train ML/DL models. The software team will develop a system to allow dental practitioners to apply different trained ML mechanisms to new images for the detection of patient illnesses. This will entail creating the ML framework; allowing image collection, cataloging, and selection; providing flexible capabilities to incorporation of new analysis methods and lesion/illness types.
more detailNeuroSymp is an AI-driven platform designed to evaluate symptoms associated with common neurological disorders offering early detection, medical care recommendations, and educational resources. Targeting individuals experiencing neurological symptoms I hope to bridge the gap between symptom onset and diagnosis by focusing exclusively on neurological conditions like migraines, epilepsy, Parkinson’s disease, and multiple sclerosis. With features like guided symptom questionnaires, urgency-based care guidance, and a user friendly interface, NeuroSymp can provide a symptom analysis and immediate results. Its impact lies in improving timely care by reducing diagnostic delays therefore enhancing outcomes for patients with neurological concerns.
This project consists of developing an image classification based algorithm trained on a wide skin color range of dermatoscope (an imaging instrument used in clinics, which will allow the images to be standardized) images of skin diseases to accurately identify skin cancer. The goal of this project will be to improve the accuracy at which skin cancer is detected in darker skin tones, specifically in the beginning stages. Serious skin cancers such as melanoma have a five-year survival of 97% in earlier stages as compared to 14% in later stages (StanfordReport). In particular, visual examination with a dermatoscope is used to determine the presence of a skin disease from a mole, and then a physician can proceed to classify the skin cancer.
The process of building the deep neural network, while a valiant goal, can be shortened by condensing the project to simpler image classification with specific parameters. Katie Heath, a BME PhD student at UNC, has developed an algorithm with a similar purpose for cancerous tumors in the brain. The results will be similar to what was achieved by an algorithm developed by Google trained to identify 1.28 million images from 1,000 object categories (GoogleNet Inception V3 CNN architecture). This algorithm has been trained to differentiate between cats and dogs, but it can be trained to identify and classify skin abnormalities, with an emphasis on skin cancers. The Stanford study attached below provides background on the type of results needed and the potential flexibillty and scope of the project. Dermatoscope images needed to train the algorithm can be secured from an online ISIC database or through UNC Hospitals with appropriate HIPAA guidelines followed. Data will be anonymous. Images will be classified and sorted into categories, with an emphasis on gathering large amounts of data from darker skinned patients, and images can be inputted as pixels with data labels. Biopsy confirmed images will be used to train the algorithm.
Upon development of this algorithm, the results can be analyzed using a sensitivity-specificity curve, which measures the algorithm’s ability to accurately identify malignant and benign lesions. This algorithm will be built as a computer application that uses data from dermatoscope images during routine skin checks at a dermatologist or primary care clinic. Our ultimate goal for this project is to analyze and improve the accuracy at which these algorithms can detect malignant and benign skin cancers on darker skin tones, using standardized dermatoscope images to facilitate easier completion of the project. This will be the first step in eventually creating an explainable AI powered application that doctors can reliably use to supplement and enhance their diagnoses.
AI in Melanoma Stanford Study Link:
As a new college student, I have been tasked with the undoubtedly difficult job of managing almost all facets of my life. Of course, this sounds silly at first, but it is an obstacle that almost all can relate to. For some, it may come earlier or later in life, but the same realization appears nonetheless. That realization being that you have absolute responsibility of managing your own life.
Now, food is one of the most, if not the most important part of your life. I mean, we literally can’t live without it. Your thoughts linger around the question of where you’re going to get your food from, but the answer to that question is not the end of the journey. You must then ask if that food can even be eaten. And I don’t just mean can I viably eat the probably undercooked ramen I made in the corner microwave of the kitchen on my dorm floor. I mean, am I allergic?
It can be hard for a person still learning how to balance the rest of the responsibilities in their life to also correctly manage such an important question. That’s where SafeEats comes in. For the customer trying to manage their food allergies, they simply enter in their known food allergies, and we handle the rest. After a customer enters their allergies, the allergies are cross-referenced with the menus of restaurants to give the customer an easily accessible list of the foods on the menu which they can eat.
SafeEats would be a mobile application that served both iOS and Android devices. Since the app serves as a way to make dining easier for the customer, it would have links to other applications such as DoorDash, Postmates, or the actual online ordering system for the restaurant, and potentially a partnership with any of the previously named organizations.
Because these companies have already mastered online ordering, SafeEats would be unsustainable if it were to have online ordering of its own. Ideally, the application only makes the items on the menu which the customer can actually eat available to them, so as to avoid confusion or any possibility that they choose an item they are allergic to.
Approximately 32 million people in the United States are affected by food allergies, so the problem this app is solving is simple, dining accessibility for those with food allergies. Although I used a college student as an example of someone who would greatly benefit from this app, another relevant demographic is parents, as the safety of their children is of the utmost importance. While dealing with the pressures of raising a child, they would absolutely benefit from an app that ensured the safety of their children when eating food from restaurants. Overall, the app has one goal and one goal alone, easy dining in a safe way.
This is a continuation of an earlier COMP 523 project (S'24).
SafeEats is a mobile application with a format similar to that of DoorDash or GrubHub, and will be available for both iOS and Android devices. Around the world, food allergies have become more widespread and more severe, with hospital visits stemming from food allergies increasing threefold between 1993 and 2006, and hospitals seeing the largest amounts of cases in history. SafeEats serves as a way to make dining out more accessible and safe for those with food allergies. After downloading the app, the user enters their food allergies into the application's database, saving this to their client profile for reference. The application then cross-references the user's food allergies with the ingredients on the menus of restaurants and fast food chains, showing the user the menu items which they can eat. Therefore, the issue of safe online ordering and easier accessibility for those with food allergies (and parents /guardians of children with food allergies) are solved through the application. There are not any other particular systems which SafeEats must interact with. Supporting Doc