Eric C
- Research Program Mentor
PhD candidate at University of Michigan - Ann Arbor
Expertise
Medical Imaging, Computer Vision and Object Tracking, Machine Learning, Deep Learning, Image Processing, Natural Language Processing, Large Language Models, Generative AI
Bio
Hello! My name is Eric, and I am currently a Ph.D. Precandidate at the University of Michigan - Ann Arbor. My research involves optimization techniques and machine learning for medical imaging applications, primarily Single Photon Emission Computed Tomography (SPECT). I am also an Algorithm Engineer at KLA, where I leverage deep learning techniques to measure structures on silicon wafers that are only a few nanometers large! I have been a mentor with Polygence for over a year! In my free time, I enjoy weightlifting, reading, video games, coding, and spending time outdoors with my two beloved dogs. I was able to hit my lifetime goal of a 500-pound deadlift late last year! I'm working now on being more mindful and learning to be present every day. I am thankful to be able to add Polygence mentor to my list of current activities and am looking forward to helping you reach your project goals :).Project ideas
Deep Learning for Automatic Brain Tumor, Edema, and Necrosis detection
For this project, the goal is to create, train, and deploy a deep neural network on images of MRI brain scans to look for a series of dangerous brain conditions. By creating this network, we can ensure that each patient scan receives two examinations; one from a radiologist and another from our product. Doing so reduces the odds that any harmful condition may be missed.
Automatic Bird Identification using Deep Neural Networks
Imagine you were a conservationist who wanted to track the presence of specific types of birds in a given area. However, the birds are nocturnal or extremely rare, so they can hardly be caught by cameras. This project would enable wildlife researchers to identify bird species purely based sounds they make while chirping! Using deep learning, the goal of this project is to create a model which can accurately identify bird species based on audio signals alone.
Help catch bad guys, with image filtering!
One of the most common tasks in image processing is image denoising. From pictures on your cell phone to the notoriously low-quality images of gas station security cameras, a noise-free image is a desirable image. In this project, one will learn the basics of denoising images using a variety of modern image processing techniques for their desired imaging modality. This could be medical imaging, optical imaging, or even radar imaging!