REU 2024

Born and raised in Cuba, my journey into the captivating world of mathematics began at a young age, igniting a passion that has guided me through numerous challenges and triumphs. My venture into mathematics took a significant turn at the age of 12 when I participated in the Cuban Mathematical Olympiad. Despite facing setbacks and years of competition without victory, I persevered. Achieving a perfect score in the state competition during my freshman year of high school renewed my confidence. Although narrowly missing the top 10 at the National Olympiad was disheartening, I refused to give up. The pandemic provided an unexpected opportunity for me to deepen my studies. However, my hopes were dashed when the Olympiad was canceled upon my return to school. With just one shot left, I poured my heart into the competition, ultimately earning second place nationally in the 2022 National Mathematical Olympiad. Throughout my academic journey, my passion for mathematics remained unwavering, even amidst challenging living conditions at Martires de Humboldt 7 High School. As a boarding school student, I found solace in pursuing my academic goals, fueled by a desire to give back to my community. Organizing mathematics sessions for fellow students solidified my passion for teaching and mentorship. Upon immigrating to the United States, I dedicated myself to mastering English and furthering my academic pursuits. As a tutor at Miami Dade College, I found fulfillment in helping others while continuing to expand my own knowledge. I have also continued to promote recreational mathematics in Miami Dade College through the Mu Alpha Theta Honor Society. Recently, I was offered the opportunity to spend the summer conducting research in computer vision at the University of Central Florida (UCF). I am thrilled about this prospect, as it allows me to delve deeper into a cutting-edge field and apply my mathematical skills to real-world problems. This experience will not only enhance my knowledge and expertise but also bring me one step closer to achieving my long-term goals. I am eager to embrace this new challenge and make the most of this incredible opportunity.

Hello! My name is Genesis Escobar, and I am a senior at the University of Central Florida and am originally from Saint Augustine, Florida. I am majoring in Computer Science, and I have taken previous coursework related to Computer Vision such as Robot Vision and Artificial Intelligence. I also have math coursework such as Statistics l & ll, Calculus l & ll, and Discrete Structures ll. I was interested in joining this REU to build on my knowledge of computer vision with a research project and gain valuable experience working with mentors. I have learned so much about the research process so far! I am working with Dr. Chen Chen and Dr. Yunjun Xu on the “Geolocation of Strawberries”. I will create a dataset of strawberry coordinates that takes into account different lighting and height from the camera to the strawberry. With this dataset, I will train a deep learning model. Then, calibrate the camera to map the strawberry coordinates and achieve image segmentation object detection. For any questions about this project, this is my contact: gescobar2002@gmail.com.

My name is Joseph Ho and am a rising sophomore at the University of Wisconsin - Madison double majoring in Computer Science & Data Science. I’ve had previous experience in programming, as I’ve taken multiple courses in various CS topics and done research in deep learning in medical imaging. As someone who’s extremely interested in the world of machine learning, this REU program has provided me an excellent opportunity to learn and gain experience in computer vision. This summer, I’m working with graduate student Nyle Siddiqui and Dr. Yue Wang on investigating and probing diffusion model by slightly editing their forward and reverse processes for various applications. Some applications include differentiating real images to diffusion-generated images, and detecting if images were memorized from training data. Feel free to reach out to me for further questions at hojoseph2006@gmail.com!

Hi everyone! I'm Jevon (Jay) Joseph, and I'm currently a rising junior at Randolph College (I don't know where that is either). I'm a computer science and math major with a minor in global studies. This REU has been my first introduction to ML and CV, but I've already learned so much here! I've currently taken: computer programming I & II, calculus 1 & 2, data structures, operating systems, algorithms, and I've had previous intern experience at a coding camp. My current research involves Time-of capture predictions. For this project, I'll be working with David Shatwell, Kevin Zhai, and Dr. Mubarak Shah to enhance video metadata, aid in digital forensics, and improve retrieval systems using various outdoor factors such as: light intensity, shadow placement, seasonal weather conditions, and foliage. For this, we'll need well calibrated outdoor videos, rich in temporal information to predict the hour and month that a given video was taken. For updates about my journey here at UCF, please look at my weekly presentations and more below! Don't hesitate to contact me at jjoseph12312@gmail.com Bye bye 🙂

My name is James Moseley, I am an aerospace engineering major at UCF. I was granted the opportunity to join this research team by Professor Niels Lobo, I was excited to face the challenges of machine learning. I have taken multiple CS and Calculus classes which helped as a framework for the more advanced topics we are currently researching. My current research topic is self-supervised retrieval of geolocation using a transformer.

Hello there! I am a rising senior at Northeastern State University, located in Oklahoma. I am currently majoring in computer science with a minor in mathematics. This is my first REU experience, and it is an honor to have been chosen to research computer vision in such a prestigious program as UCF! I am working with Dr. Mubarak Shah and James Beetham on evaluating and bypassing data-free model stealing defenses, which includes previously researched model defense methods and determining the effectiveness of the defense by using modified model stealing attacks. We will also assess this new method on an undefended model to see the results. Feel free to review our progress at the links below! If you have any questions, you are welcome to reach out to me at brett3000@gmail.com.

I am currently a junior pursuing a dual major in Computer Science and Physics, complemented by a minor in Mathematics. My academic journey has fostered a deep interest in computer vision and artificial intelligence, driving my passion for innovative research in these fields. I am thrilled to embark on my latest research project, "Synthetic Meets Real: A Hybrid Approach to Boost Clip Model Performance," under the guidance of Adeel Yousaf and Dr. Mubarak Shah. My previous experiences include attending UCF's annual hackathon, KnightHacks, where I collaborated on various personal projects that honed my technical skills and fueled my enthusiasm for problem-solving through technology. As I delve into this REU project, I am eager to contribute to advancements in computer vision and explore the intricate interplay between synthetic and real-world data. If you have any questions or would like to discuss my research more, you can contact me at aisiddiqi05@gmail.com.

My name is Jasmine Sun, and I am a rising sophomore at Washington University in St. Louis majoring in Computer Science + Mathematics and Economics. I have had some experience with programming and other CS topics, but this REU is my first time studying computer vision. This summer, I am working with Dr. Laura Brattain on Uncertainty Estimation in Medical Image Deep Learning. To do this, we are using neck ultrasound images to develop a deep learning bounding box detector using YOLOv9, which will be trained to identify and classify key anatomical landmarks in the neck that are used for emergency airway access. With this, we are quantifying and reporting the model’s epistemic uncertainty in its recognition. To estimate uncertainty, we will use a Gaussian Distribution function to calculate the likelihood that a detected object belongs to its known class. Feel free to reach out with any questions at j.h.sun@wustl.edu.

Hi, I’m Keerthi Veeramachaneni, a rising 3rd year student at the Georgia Institute of Technology. This REU program is my first exposure to computer vision and I’ve had a great time learning more about it. I’m currently working with Praveen Tirupattur and Dr. Mubarak Shah on the AI-Generated Video Contect Detection using Vision Language Models project. We’re working on the creation of a model to distinguish synthetic images from real images, and developing a corresponding dataset. For more details, check out the materials below and feel free to contact me at aveerama3@gatech.edu with any questions!

My name is Claire Zhang, and I am a rising junior at Brown University studying Applied Mathematics-Computer Science. Although I have taken several computer science courses prior to this program, the University of Central Florida’s REU Program in Computer Vision is my first exposure to Computer Vision. This summer, I am working with Max Liggett, Mengjie Li, and Kristopher O. Davis on SEM Image Analysis for Material Characterization. In this project, we are working to develop an automated segmentation model that helps identify corrosion of photovoltaic cells, indicated by darkened sections in cross-sectional Scanning Electron Microscopy (SEM) images. Our approach is to use a U-net architecture to identify materials in these images (e.g. glass) and extract information on pixel brightness and spatial information. If you have any questions, please feel welcome to email me at claire_y_zhang@brown.edu.