Hi there!—안녕하세요!

I am a generalist with diverse interests—and love delving into unfamiliar fields within engineering.

Background

My areas of expertise-ish are:

  • Perception, control, and motion planning for robotic systems
  • Inference optimization of machine learning models (for edge inference)
  • Building software in general (web, mobile, desktop, backend/database, systems-level)

My research interests are:

  • Applying computational approaches to new domains
    • Including traditional sciences, unobservable or subjective phenomena, and more
  • Utilizing joint information representations and test-time input augmentations to extend the capabilities of large transformer models (e.g., LLMs)
  • Devising human-centered communication and control schemes by considering human perceptual, cognitive, and motor characteristics
  • Utilizing large-scale unlabeled data (e.g., videos) for weak self-supervised learning of joint/multimodal information representations
    • Including value functions, world-state perception and prediction, and more

Vision

My life's goal is to tackle "underserved" problems and create solutions that directly improve individuals’ quality of life.

So far, I have encountered a few important topics that fit this description:

  • Access to high-quality knowledge content for non-English speakers
    • An unfair amount of academic lectures and articles are available strictly in English.
    • As a society, we have powerful machine-translation technology, but it has not yet been deployed in end-user tools that let people comfortably translate textbooks, lectures, and assignments—and learn in their own language.
  • Assistive technologies that enhance life for people with sensory impairments
    • The world we live in is not designed with sensory impairment in mind. This includes roads, products, interfaces, schools, and public buildings. Our society presents people with more challenges and safety risks while offering them fewer opportunities for success in school and work.
    • We have lots of room for improvement as a society and as a field. Some of the work is social and infrastructural, but modern technology also has a great deal to offer in reducing the frustrations that arise during activities of daily living.
    • Visual impairment, for example, is both more common and more detrimental than many people realize. I am not currently putting as much time into visual-assistance systems as I once did, but I plan to continue until the problem is truly considered “solved.” Projects like Project Astra are very exciting!

I plan to keep working hard on things like this throughout my life.


Talk

I am always open to chat. Send me an email at: