UROP Research Mentor Project Submission Portal: Submission #1187

Research Mentor Information
---------------------------
Primary Research Mentor Name: Jonathan Adams
Research Mentor Preferred Pronouns: {Empty}
When potential research assistants are reaching out via email, what is your preferred honorific?: Dr.
Contact Email (FSU Email if affiliated): jladams@fsu.edu
Position Title: Faculty
FSU College (if applicable): Communication and Information
FSU Department or Non-FSU Organization Affiliation: Information
Headshot (optional): https://cre.fsu.edu/system/files/webform/urop_project_proposal_portal/20531/headshot.jpg

Additional Research Mentor(s)
-----------------------------
Research Assistant Supervisor (if different from above): Benhur Ravuri
Research Assistant Supervisor Preferred Pronouns: {Empty}
Research Assistant Supervisor Preferred Honorific?: {Empty}
Contact Email (FSU Email if affiliated): bravuri@fsu.edu
Name of Other Faculty/Collaborator(s) (if applicable): Odin Moja
Other Faculty/Collaborator(s) Preferred Pronouns: {Empty}
Other Faculty/Collaborator(s) Preferred Honorific?: {Empty}
Contact Email (FSU Email if affiliated): omoja@fsu.edu

Overall Project Details
-----------------------
Title of the Project: AI Agents (Agentic AI), Ocean Search and Rescue (computer vision), AI Retrieval Augmented Generation
Project Keywords: MLab, computer vision, AI Agents, RAG
Are you currently looking for research assistants?: No
Number of Research Assistants Needed : 3
Relevant Research Assistant Major(s):
Open to all majors, but we prefer students who have an ability to code (no 'vibe' coders please), interest in computer systems, artificial intelligence, and publishing basic research. 
Email your resume to jladams@fsu.edu. I will respond with a meeting time on the 2nd, 3rd, or 4th of September.



Project Location:: On FSU Main Campus
If the project location is off campus, does the research assistant(s) need to provide their own transportation?: Yes
Please select the choice that most accurately describes your expectations for the research assistant(s):: In-person
Approximately how many hours a week would the research assistant(s) need to work?: 6-10
Roughly what time frame do you expect research assistant(s) to work?: Flexible schedule (Combination of business and outside of business. TBD between student and research mentor.)
Overall Research Project Description:
The MLab research specializes in technical research that uses synthetic media to train AI. Our research has explored various ways to use "deep-fakes" of subjects that are hard to photograph or in cases where photos are not sufficient in number (this is usually the case). We developed tools to generate labeled images ready for training. This opens up a wide variety of opportunities to explore synthetic media: AI training as well as object, and anomaly detection.

Should you decide to apply, and ultimately participate, the lab hosts undergraduate alumni of the lab, graduate students and professors who can assist you with your chosen project. Interested students should understand that participating in the lab means you are a self-starter who works well in small groups of like-minded students.

Working knowledge of Python (or other language) programming language is necessary, the projects are intended to get MLabbers up to speed. One of the primary missions of the lab is to encourage research related to the use of Artificial Intelligence agents, and supporting students who desire to publish their works. Since the MLab was founded in 2019, the MLab has produced 13 student-led research papers that have been accepted by international conferences.

In 2023 the MLab had a grant that led to the development of a Binaural Beat Box, which we proposed to use as a Palliative care tool. The Beat Box uses generated synthetic music, Art, and Animation for a tablet application. 

Other potential areas of research
Music generation, music transcription, segmentation
Retrival Augmented Generation
Computer vision/synthertic media
Agentic AI 
  


Research Tasks:
Research tasks include a wide variety of activities. Planning, literature reviews, data collection and analysis, 3D model generation, 2D image generation, working in teams.
MLabbers work in teams to set up an algorithm, learn how to train the algorithm and use the resulting work to address a real problem. Specifically, working knowledge of the Python programming language is essential. Prior experience with another programming language such as Java or C++ are acceptable. We use Anaconda Navigator in the lab, students receive instruction on how to work in a lab, using conda environments, how to set up a work environment, use GitHub and copilot. 
Each year, UROP students publish at least one article in a conference. The MLab has a steady output of articles published in respected IEEE and education conferences. UROP students, then learn how to conduct a literature review, establish a research agenda, work through the research project to completion. Coaching is provided to turn the resulting work into a publication. 


Skills that research assistant(s) may need::
Required:
Ability to work collaboratively.
Recommended:
Coders, CS, Engineering, Statistics, and 3D artists who can model different subjects. 


Mentoring Philosophy:
Teamwork and collaboration are fundamental to our lab's success. To illustrate this, consider the Peloton - a group of professional cyclists working in perfect coordination. This fascinating display of teamwork requires cyclists to take turns leading the pack, each giving their maximum effort at the front until exhaustion sets in. When the lead rider can't continue, they fall back to the rear of the group, while another cyclist moves forward to take the lead position.

The Peloton creates a wind vortex (draft) at the rear of the group that powers collective action. As riders retreat from the lead position, they can rest/recover while being pulled along by this draft, which requires minimal effort. This rotation continues, with each team member taking their turn at the front before falling back to recover.

This collaborative approach illustrates why even the strongest individual cyclist cannot outperform a well-coordinated Peloton. Similarly, in the MLab, our strength lies in our collective effort. Each member contributes by alternating between leading projects and supporting others' initiatives, fostering a culture of mutual respect and collaborative achievement.
As your mentor, I'm here to guide and support your journey. The MLab provides a safe environment where trial and error are not just accepted but expected. You're encouraged to take risks, learn from mistakes, and grow through experience, knowing that guidance and assistance are always available. My role includes serving as a guide to introduce MLabbers to research and cutting edge technoogy.


Please provide a link to your publications, a video clip, or a website for your research project (if applicable):: https://directory.cci.fsu.edu/jonathan-adams/
Please add any additional information here (if applicable)::
Equipment
The MLab has 5 Machine Learning stations.
1 Cyberpower 2080 Ti, 64 gb CPU RAM
2 Alienware 3090, Cyberpower 3090 64 gb CPU RAM
2 Alienware 4090, 64 gb CPU RAM 
Oculus Headset
Fixed wing and rotary drone aircraft.



Are you interested in participating in the UROP Research Mentor Roundtable?: Yes
Roundtable times and Zoom links:
{Empty}

Roundtable Info:
- Day: Tuesday, September 2
  Start Time: 12:00
  End Time: 3:00
  Zoom Link: https://fsu.zoom.us/j/99871326598
- Day: Wednesday, September 3
  Start Time: 12:00
  End Time: 3:00
  Zoom Link: https://fsu.zoom.us/j/99871326598
- Day: Thursday, September 4
  Start Time: 12:00
  End Time: 3:00
  Zoom Link: https://fsu.zoom.us/j/99871326598

UROP Program Elements
---------------------
Mentor Handbook, FAQs, and Communication: Yes
UROP Performance Evaluation: Yes
Materials Grant: Yes
UROP Poster Presentation: Yes

{Empty}
Year: 2025
update url: https://cre.fsu.edu/urop-research-mentor-project-submission-portal?element_parents=elements/research_mentor_information/headshot_optional_&ajax_form=1&_wrapper_format=drupal_ajax&token=LtS5DGNcTsXlD9R3IEpWW9xlTYFJX5ZZazInvMZNAB8