M

i

r

u

J

u

n

M

i

r

u

J

u

n

    [ TL;DR ]

  • Data Scientist @ US Naval Base
  • B.S. in CS @ University of Southern California (USC)

    • & Minor in 3D-Animation

    2024, Summa Cum Laude

  • T3 Secret Clearance

    • Excited about building intuitive and efficient human-computer collaboration through AI in XR and robotics :)

[ PROJS ]

Estuary: AR Companion Project

WebsiteLonger DemoCHI25 -- Interviews

Framework with client-server architecture on Quest 3 and Apple Vision Pro

Memory Space: Chinatown AR

Website/WebAR LinkLab WebsiteGrand Prize Article

A geolocated WebAR experience on Niantic's 8th Wall as part of a temporary exhibition in LA's Union Station.

NASA SUITS Challenge 2022-2023

2023 USC Article2022 USC Video

An Augmented Reality application on the Microsoft HoloLens2 (with MRTK2) to help astronauts with navigation, localization, terrain mapping, biometric and telemetry data, UIA panel instructions in low-light conditions.

This project evolved into the Estuary project!

Trajectory Improvement with Comparative Natural Language Feedback

WebsiteCoRL'24 PaperHIRL'24 Paper

Using comparative feedback in natural language to improve a robot arm (WidowX) trajectory.

[ SIDE QUESTS ]

[ Remembering I have free will in an open-world environment ]

Cringe Canon

GitHubHackathon

Takes your cringe, makes it canon.

Upload pictures of your original character (OC) art and generate "canon" character profiles and images. Chat with your character while sending camera snapshots of your world!

Sketch Artist

GitHub

Local image generation and in-painting pipeline via CLI.

Containerized; meant to run on NVIDIA Jetson Orin Nano.

QueueTify

GitHub

Playlist name generator with an asynchronous task queue and results database.

Containerized; locally powered by llama3.2, through Ollama.

To The Clouds

360 VR short film for class (CTAN 504).

Made in Maya. Largely made for viewing on a Quest 2.

[ PUBS ]

  • S. Lin, M. Jun*, B. Rizk*, K. Shieh, S. Fisher, and S. Mozgai, “Optimizing SIA Development: A Case Study in User- Centered Design for Estuary, a Multimodal Socially Interactive Agent Framework,” in ACM Conference on Human Factors in Computing Systems (CHI), 2025.
    • arXivVideo
  • S. Lin*, B. Rizk*, M. Jun*, A. Artze, C. Sullivan, S. Mozgai, and S. Fisher, “Estuary: A Framework For Building Multimodal Low-Latency Real-Time Socially Interactive Agents,” in 24th ACM International Conference on Intelligent Virtual Agents (IVA), Dec. 2024.
    • arXivWebsiteLinkedIn
  • Z. Yang, M. Jun, J. Tien, S. J. Russell, A. Dragan, and E. Bıyık, “Trajectory Improvement and Reward Learning from Comparative Language Feedback,” in 8th Annual Conference on Robot Learning (CoRL), Sep. 2024.
    • arXivWebsiteVideo
  • J. Tien*, Z. Yang*, M. Jun, S. J. Russell, A. Dragan, and E. Bıyık, “Optimizing Robot Behavior Via Comparative Language Feedback,” in 3rd Human Robot Interaction (HRI) Workshop on Human-Interactive Robot Learning (HIRL), 2024.
    • PaperVideo
  • M. Potter*, and M. Jun*, “Do Bayesian Neural Networks Improve Weapon System Predictive Maintenance?,” in Annual Reliability and Maintainability Symposium (RAMS), Jan. 2024.
    • arXiv