REU 2019

I am a rising senior at Xavier University, and I will be completing my B.S. in Computer Science and Mathematics with a minor in Statistics this upcoming spring. After graduating, I plan to pursue a PhD in Computer Science so I can work as a professor afterwards. This summer, I am working with Dr. Yogesh Rawat on cross-view action synthesis. The goal of this project is to take an action video and transform the video as if it were taken from a different viewpoint. This project has little prior research, so we are looking to set a baseline for what can be accomplished in this task. To accomplish this, we use the input video as well as a single image of the desired view, which serves as appearance conditioning. This appearance conditioning allows us to generate the video from another view even if the background or the person changes. Our progress can be seen in my weekly presentations. Feel free to contact me at schatzk@xavier.edu for more information

I am a Junior at North Carolina State University studying Computer Science. My interests include machine learning, AI, and computer vision. I chose to participate in this program because I wanted to learn more about computer vision and get a glimpse of what computer science research looks like. I am working with Amir Mazaheri on an Automatic Deep Image Cropper and Enhancer, where we try to extract an image from a photo of the image. Our goals for the project have been to design a model that could effectively crop and enhance a given image, to try training such models on synthetic datasets, and to explore what levels of image processing is possible with such a network. My email is apott@ncsu.edu if you would like more information about our work.

I am a senior studying Computer Science at the University of Puerto Rico at Bayamon. The reason that I wanted to participate in this REU program at UCF is to experience research and development of deep learning applications in the field of Computer Vision. This summer, I am working with Jyoti Kini on Weak-supervision based Multi-Object Tracking. The goal of the project is to design a MOT system in which tracks objects by finding dense correspondence between a set of images. Feel free to contact me at alex.ruiz3@upr.edu for more information.

I am a rising junior at University of New Mexico in Albuquerque, New Mexico doing double degree in Computer Science and Studio Art. This summer, I am working with Dr. Mubarak Shah and Aidean Sharghi on Video Caption Generation. In this project, we would like to automatically translate all possible events in a video into natural language sentences using LSTM-based encoder and decoder. Feel free to contact me at nta@unm.edu for more information.

I am a senior at the University of Texas at Dallas finishing my B.S. and beginning my M.S. in Computer Science. I was interested in this program because I have wanted to work with machine learning and computer vision as well as experience what research is like. I have been working with Praveen Tirupattur on a project about Real-Time Video Anomaly Detection. The main goals for the project included exploring different approaches used for anomaly detection, building a model for spatiotemporal detection, and applying it to a dataset that represents real-world anomalies. Our progress can be seen in the presentations below. I can be reached at urving4@gmail.com for more information. 

I am going into my third of five years in the Stonehill College and University of Notre Dame 3-2 Computer Science and Engineering Program. I have a strong interest in the fields of philosophy, psychology and computer science, which have a natural intersection in the field of machine learning. This is precisely the reason I chose to take part in UCF’s Research Experience for Undergraduates program funded by the National Science Foundation. I decided to work with Aayush Rana and Dr. Yogesh Rawat on Actor-Object Relation in Videos, as it was a fascinating problem which encompasses a basic form of reasoning in discerning between subjects and objects in actions and relations. Our approach is different from many state-of-the-art approaches, as it does not rely on region proposals, making it truly end-to-end. Though we encountered many challenges with the dataset, we were able to achieve some interesting results. For more information, feel free to contact me at vbobyr@students.stonehill.edu.

I am incoming junior working on my Bachelors in Computer Science in the University of Central Florida. The 2019 CRCV REU has been a tremendous learning experience. I had little prior experience in Computer Vision and Machine Learning, and this program has helped further my understanding of these fields. I am interested in both Computer Vision and Natural Language Processing, so working on Visual Question Answering with Aisha Urooj Khan was a great way to explore these areas of research. Our project focused on investigating the ways which Long Short Term Memory architectures (LSTMs) look at video, questions, answers, and other features of our dataset and make predicted answers. The goal of this project is to determine how attention-based techniques affect what these LSTMs focus on when making their prediction. Look at the links below for a fuller understanding of the project. For more information, feel free to contact me at achonculada@knights.ucf.edu.

I am a senior pursuing a B.S. in Computer Engineering at the University of Central Florida. I chose to participate in this program to gain knowledge and explore the fields of AI and machine learning. I hope to apply this experience along with the computer engineering to help advance the research community. I worked with Dr. Mahalanobis and Babak in moving target detection using infrared sensors during the REU program this summer. The research is about creating a network to detect targets that are more than 4km away. I chose to participate in their research because I was intrigued by the capability of computers to detect objects that are very difficult to identify. By helping this research, I sought to make computers be able to do more than what we can do with our own eyes and distinguish much difficult identities in videos. Feel free to contact me at arkitagishi@gmail.com for more information.

I am an incoming junior at UCF studying Computer Engineering (CpE). I have been working with AI and machine learning since last year with AI@UCF, a club that specializes in teaching people about AI with little to no background experience. With this and experience from online courses, I joined the CRCV department as an undergraduate researcher in fall of 2018. AI is my passion, and participating in the CRCV REU helped me really focus on learning more and applying my skills to a difficult problem. I started on a new project for the REU: Multi-UAV to UAV tracking. The problem of being able to track other UAVs from a UAV mounted camera is starting to come up with the increased use of UAVs in personal, business, and government applications. A fast, efficient, and accurate system for classifying small fast moving objects is needed to make a robust collision avoidance system for these UAVs, who have limited computing power and room for additional sensors. My approach is using a full 2D convolutional network that predicts pixel-wise segmentation based on temporal and spatial information for detection of the UAVs in these videos. Currently I am working to develop a novel solution for tracking the detected UAVs. If you have any questions and would like to contact me, you can reach me at brandon.silva209@knights.ucf.edu.

I am a junior at University of Central Florida majoring in computer science. I returned to school for my second bachelors degree after getting a masters degree in industrial and organizational psychology and spending a year working in business analytics. My interests are in machine learning, computer vision, and AI. I felt that this was a great opportunity to get hands on experience in research and to improve my programming skills and knowledge in deep learning. I am working with Kevin Duarte on weakly supervised video object segmentation. Our goal for the project is to segment an object throughout a video after given only a scribble of the object in the first frame. My email is danny.cosentino12@knights.ucf.edu if you would like any information about the research.