THOTH: Task-Aware Semantic Memory for Aerial Missions Joseph Letobar, Arianna Lee, Jylen Tate, Neal Cronin, Clint Johnson
MIT Undergraduate Technology Research Conference (URTC), 2026 (accepted)
research paper
THOTH (Task-aware, Hierarchical, Open-vocabulary, Traceable Heatmap),
a task-aware semantic memory system for long-horizon aerial missions.
Rational Episodic Memory: When Should Wearable AI Remember? Joseph Letobar European Conference on Computer Vision (ECCV) Wearable AI Workshop, 2026 (accepted)
research paper Also presented as an extended abstract at the ECCV Human-Inspired Computer Vision Workshop: research paper.
Engagement and surprise, respectively inspired by Theory of Mind research and predictive processing, serve as cognitively inspired signals for deciding what wearable AI should remember.
Plug and Play Myoelectric Control Joseph Letobar, Mohsen Rakhshan
Evaluating the effects of force control on the intuitiveness of myoelectric prosthetic hand control.
Augmented Reality Bowling Joseph Letobar Personal project, 2026
source code
Real-time lane and ball analytics using computer vision.
Miscellanea
Work Experience
Sep 2026 - Present, Orlando, FL — Center for Research in Computer Vision (CRCV), Undergraduate Researcher
Jul 2025 - Present, Orlando, FL — Laboratory for Interaction of Machine and Brain, Undergraduate Researcher
May 2026 - Aug 2026, Atlanta, GA — Georgia Tech Research Institute, Research Intern
Dec 2025 - May 2026, Orlando, FL — Lockheed Martin, Quality Engineer (Part time)
Apr 2025 - Jul 2025, Orlando, FL — UCF Information Technology, Information Technology Intern
Education
Expected Aug 2028, Orlando, FL — University of Central Florida, B.S. in Computer Science