URA

An Undergraduate Research Assistantships

See here for details.
I want to try my hand at research and learn about cool things in a fundamental way.

Cool Opportunities That I Am Interested in

•  Software and Hardware for Autonomous Car Racing with Yash Vardhan Pant
• Deep learning-based transfer learning models for real-world pose estimation with Sirisha Rambhatla
• Applied Computer Vision (sports analytics :)) with David Clausi
• Intelligent Connectivity for the Internet-of-Things. with Omid Abari
• Machine Learning Tools Software Development with Mark Crowley
• Code Quality with Dietl Werner
• STARMX with Ladan Tahvildari
• 3D graphics web development with Olga Vechtomova
• Development of web-based application for a deep learning NLP project with Olga Vechtomova

My tops picks are the URAs with:

• Mark Crowley
• David Clausi
• Olga Vechtomova
• Dietl Werner
• Yash Pant

Waiting

I emailed the five professors I chose from and I’m now waiting on more details about the URA and responses!

Update: I heard back from 2 profs but then heard nothing… I should reach out again and start applying to more URAs.

Next Email List
In order reach out to…

• Sirisha Rambhatla (Deep learning-based transfer learning models for real-world pose estimation) ✅ 2024-11-08
• Wojciech Golab (In - Memory Data Structures for Next Generation Multi-Core Machines) ✅ 2024-11-29
• Guang Gong (Implementation of Universal Password-less Authentication) ✅ 2024-11-08
• Sebastian Fischmeister (Improving experimentation in Computer Science and Computer Engineering) ✅ 2024-11-08
• Arie Gurfinkel (Automated Software Verification and Program Analysis) ✅ 2024-11-08
• Vincent Gaudet (Software simulation of decoding algorithms to decode low-density parity check codes) ✅ 2024-11-08

URA Results

I met with Olga Vechtomova and Moojan Ghafurian about URAs in deep learning surrounding artistic and social robotic applications.
I am leaning towards a URA with Prof. Ghafurian where the application would be developing scenarios for two social robots used as tools to teach children how they should act in given scenarios. Another project she had was a robot which changed its facial features to react to the emotions displayed by a human.

The URA with Prof. Vechtomova consists of developing a ML dance coach which I imagine would involve a CV aspect as well which could be interesting.

I intend on reaching out to both professors to inquire more about these projects before making a decision. My questions are the following…

For Prof. Ghafurian
Emotion Showing Robots Q’s

• From our conversation, we concluded that for this project I would be involved with developing the ML model that decides how to manipulate the facial structure of the robot based on the sentiment of the individual interacting with the robot
• What does “being involved with the ML model” entail here? Would I be involved in both research and implementation?
• Is this role more heavily weighted towards the research side, or the implementation of any research that has already been done?
• What sort of data would I be working with here (i.e what is the data stream coming from a user)? Are we working with muti-modal data here (auditory, visual, etc)? Situation Robotics for Kids Q’s
• From our conversation, similarly I would be working on the ML model for this project as well as developing “situations” for these robots
• What does “developing situations” consist of here? Is this brainstorming potential scenarios and then implementing these scenarios on the robots through whatever controls we use (I remember you mentioning an app)?
• Where does the ML modelling aspect for this project come into play?
• Same questions apply here in terms of data stream (what is the data stream are we ingesting)?

For Prof. Vechtomova
Dance Coach

• From my memory of our conversation, one of the projects that I was interested in was a machine learning dance coach that you were developing
• I imagine my role in this project would be more so on the modelling side in both research and implementation? Would there be an opportunity to work on the front-end UI, or sensor / CV aspect of this project as well?
• What sort of data stream are we ingesting here (Is it purely visual information that we use to inform out model? Where does our training data come from? Are we using recorded streams to train our models, are you currently working with dancers to iterate on this project)

Debating between Dietls compiler research projects vs a part time job at Hevy