Olivia R
- Research Program Mentor
PhD candidate at Washington University in St. Louis
Expertise
Neuroscience, Artificial and Biological Vision, Machine Learning, Computational Neuroscience, Psychology/Behavior, Biostatistics
Bio
Hello, I'm Olivia! I'm a Ph.D. candidate in Neuroscience in the Ponce Lab at Harvard Medical School. I started my doctoral degree at the Washington University in St. Louis School of Medicine, and our lab recently joined the amazing HMS community in the Department of Neurobiology. I'm fascinated by how the brain processes vision and transforms that sensory information into essential cognitive processes, such as visual working memory. Under the mentorship of Dr. Carlos Ponce, I use a combination of classical neuroscience methods (in vivo electrophysiology), novel machine learning techniques (Generative Adversarial Networks and Convolutional Neural Networks), and innovative behavioral tasks to understand how our brains make sense of our complicated visual world. My free time is usually spent hanging out with my lazy dog, Rusty! I also love to make homemade jam and bread, tend to my garden, and knit/crochet. You can also catch me hiking and spending time in nature, as well as getting adventurous with snowboarding, cliff jumping, and traveling to new places. My personal experiences as a woman in computational neuroscience have made me passionate about increasing STEM access and inclusion to students from traditionally underrepresented groups; as a mentor, I want to pay forward the kindness and encouragement that my mentors gave to me.Project ideas
Introduction to Deep Learning and Computer Vision
Have you ever wondered how a self-driving car can identify and respond to different objects in real-time? Or how social media sites can automatically detect and differentiate between faces in photos? This project aims to give you a basic understanding of how computer vision systems work, and will culminate in you creating and training your very own neural network for object detection and classification. Prerequisites: Enthusiasm and desire to learn! Prior programming experience, especially in MATLAB, is helpful; however, it is not necessary, and you will be taught everything you need, customized to your individual level.
Comparing Artificial and Biological Visual Systems Using Feature Visualization
There are many parallels between how computer vision and biological vision systems process and categorize images, to the point that Convolutional Neural Networks (CNNs) provide a valuable model for primate object recognition. We will explore the structure and function of the primate visual system by delving into classic neuroscience literature, and then draw comparisons with a well-known CNN, AlexNet. Students will learn about how different AlexNet layers parallel different brain areas along the ventral visual stream, using feature visualization to show the student how simpler features are aggregated into a holistic image that can then be classified. Prerequisites: Enthusiasm and desire to learn! Prior MATLAB experience or knowledge of machine learning is helpful; however, it is not necessary, and you will be taught everything you need, customized to your individual level.
Comparing Saliency Prediction Models with Real Eye Movement Data
When looking at a complicated picture, certain features pop out and draw our attention; these attention-grabbing features are referred to as "salient". But how do we choose where to look first in a crowded scene? Are people generally consistent in the patterns they use to look at pictures? Can we predict salient locations based on nothing more than image properties at the pixel-level, or does saliency also depend on semantic meaning and goal-directed behavior? We will explore these questions and more, using open-source human eye movement data and several kinds of gaze prediction models. Students will learn how to predict human eye movement data using computational models, and will be able to compare these predictions with real data. Prerequisites: Prior programming experience, especially in MATLAB, is highly recommended. However, highly-motivated students with no prior coding experience, who express interest in spending additional time learning foundational MATLAB skills, are also encouraged.
Self-Directed Research Project in Neuroscience/Psychology
Maybe you're interested in better understanding the mechanisms of a neurological or psychiatric condition. Maybe you've noticed a neat pattern in human behavior that you want to explore. Or maybe you're curious about how people form new memories, become bilingual, or develop prejudices. If you find the subject of human brain and behavior fascinating, then let's chat! Through this project, students will learn how to effectively read and evaluate scientific papers, devise their own hypotheses, propose potential experiments, and write their own scientific article. Depending on the type of project selected, students may also be guided in the process of obtaining human survey responses, performing basic analyses on open-source datasets, conducting a meta-analysis in a subfield, and/or writing a scientific review article. If desired, students may choose to submit their papers to youth science journals, and will be guided through the editing, peer-review, and publication process. Prerequisites: Enthusiasm, desire to learn, and a healthy attitude towards receiving constructive feedback on their ideas.
Coding skills
MATLAB, Python, BashLanguages I know
SpanishTeaching experience
I was an Undergraduate Research Opportunity Program (UROP) mentor at my undergraduate institution, Florida State University, for two years; I led and mentored a class of more junior undergraduate students in the process of joining a lab, completing a self-directed research project, creating an APA-style poster, and presenting their research at an undergraduate research showcase. I also was a graduate TA for the undergraduate Laboratory of Neurophysiology class at Washington University in St. Louis, where I taught students about visual encoding in the horseshoe crab optic nerve and led a laboratory preparation for it.Credentials
Work experience
Education
Reviews
"Olivia is an amazing, supportive, knowledgable, and enthusiastic mentor! She exceeded my expectations by offering just the right amount of help so that I learned from my mistakes without feeling lost or overwhelmed. She answered all my questions in and out of sessions and was very flexible with scheduling. During sessions, she always had a positive attitude and encouraged me during challenging parts of my project. She asked me about my goals for the project and helped craft a plan for our 10 sessions to ensure that it was completed on time and met all my expectations. She even helped me come up with ways to expand my project. In addition to being extremely knowledgable about neuroscience, machine learning, and writing research papers, she is skilled at explaining complicated topics to high school students. Olivia was a wonderful mentor, and played a huge part in my project's success!"
