Current IVILAB undergraduate research opportunities
This page describes current research opportunities for undergraduates in the IVILAB, and how to apply. Please be aware that due to overwhelming interest in the IVILAB we need to limit the number undergraduate researchers. Unfortunately, we need to turn away many students.
General qualifications. Undergraduate research students must be able to realistically commit 9 hours a week to research. Currently we favor research for credit (i.e., honors projects or independent studies), and students in their second or third year. We have no strict general qualifications, but the following counter indications will need to be compensated by other strengths.
- Mediocre recent academic record. If your academic record is weak, you are likely best served by improving your GPA. Most undergraduates in the IVILAB have GPAs in line with that required for honors students (3.5 or above).
- New to university. If this is your first semester in university, that will likely limit the projects that you can have a positive impact on.
- Senior student. Research takes time, and a single semester is barely enough to get started. Seniors will need to address whether they are informally committing to second semesters, and why that is realistic given their load in the final semester.
If you are interested in any of the current projects described below, you will need to provide the following PDFs:
- Unofficial transcript
- Your CV
- A short statement of purpose (SOP) describing your academic history, your interest in IVILAB research and whether you are interested in credit, the specific project(s) you would like to be involved in, and relevant prior experience for that project. (If you apply for grad school, you will need to write a good SOP; consider this practice!)
Games that induce emotion for ToMCAT.
The ToMCAT project (Theory of Mind based Architecture for Teams) is about
virtual assistants for teams collaborating on missions. This part of the
project is contributing to software systems to collect synchronized data from
multiple people interacting within a Minecraft based gaming system. The initial
step here is to choose, configure, and/or build Minecraft based games that will
induce emotion in accord with the research plan. A second step will be
integrating the game into an interface to collect data and manipulate the game.
Eventually this data collection will be extensive, including video, audio,
physiologically data such as heart rate, EEG, and fNIRS, which is a state-of-the
art recording method of brain activity. There is a delay here as some of the
equipment will arrive mid October at the earliest.
This is project is a good choice for those that have some experience and interest in game development, building systems, and as well as computational understanding of affect. ToMCAT development will follow strict software development processes.
- CompTIES This project will contribute to a web interface to a program to model dynamic data. This project is well underway, and involves web development, human computer interfaces, and could include scripting and C++ programming.
- fMRI dataIn addition to the ToMCAT project that will study brain data from fNIRS data, we also work with fMRI data in the context of decision making, and also understanding Alzheimer's disease. The opportunities for undergraduate research here are less well defined at the current time, but likely will involve deep learning to predict experimental conditions and/or outcomes. Tracking shrimp and determining shrimp health. Believe it or not, our university in the desert has an aquaculture program, and we are working with researchers in that program to monitor shrimp feeding and growing. This is a very interesting project as the structure of shrimp makes using standard computer vision algorithms difficult. Typical activities will include investigating and implementing deep learning methods for adaptive background subtraction, and developing classifiers for indicators of shrimp health.
- Detecting eye-blinks from video. Eye-blinks may be indicative of dopamine levels in the brain, and thus detecting them in video can have a number of interesting applications. This project will develop and evaluate methods for doing so, beginning with finishing an implementation of a deep learning method to do so.
- Evaluating and grounding methods for detecting facial muscle movements from video. This project will involve getting existing software running, and evaluated on multiple data sets. Grounding the performance will involve collecting data in a lab equipped to monitor facial muscle movement.