We are a team of researchers and engineers who are serious in machine learning and computer vision. Here are our answers to some of frequently asked questions regarding joining our group.

Team's interest: The team's primary focus is to get you succeeded in your career path. Customizing to your interest, we try to help your way to reach the goal after you graduate the lab. Most of our team members are for academic positions (e.g., faculty, world class industry researchers) and a few are for engineering positions (e.g., world class software companies). We mostly aim for sharing our results in top venus in computer vision and machine learning (e.g., CVPR, ICCV, ECCV, NeurIPS, ICLR, ICML) and contribute to open source softwares to benefit other researchers including ourselves.
  Note that we, the team of 10, have published 15 papers in the top-tier AI venues mentioned above during past 1.5 years)

Our daily life: We want to come up with working ideas for a lot of bottlenecks in visual/multi-modal understanding problems and relevant machine learning problems. We read papers, discuss with fellow students and faculty, implement them and come up with new ideas for the issues in the state of the arts.

Open discussion: We welcome any crazy ideas to try and discuss among the team and the faculty is always open to discuss on anything.

Computing resources: We try our best to provide you the best development environments (state-of-the-art desks, noise cancelling headphones and etc.). We are equipped with more than 150 GPU's (mostly A6000); Each one will have a state of the art workstation equipped with one A6000 and a 39" wide curved monitor for fast prototyping (roughtly 8+ GPUs per person).

Monetary compensation: We try our best to give you the best of monetary compensation. In 2022, most of our team members are paid to the legally allowed maximum graduate student salaries and more for additional project wordloads if you want to be involved.

Active collaborations: We work closely with Allen Institute for AI (AI2), computer science at University of Washington, Harvard medical school via Massachusetts General Hospital, computer science at University of Minnesota, Twin City, and computer science at University of Maryland, College Park.

- Few-shot, zero-shot, continual (un-)learning: We try to mitigate the annotation cost for visual understanding and machine learning problems.

- Embodied AI: Combining few-shot, continual video understanding with language understanding, we try to architect new models to build a robotics agent to help household tasks (e.g., bring a cup of water from the kitchen).

- Multi-modal AI: We try to build models to understand languages alongside with visual signals. Other than the vision-and-language understanding, we are also interested in various modalities including sketch, diagram and neuromorphic (or event) understanding.

- Video understanding: We try to architect new models for understanding videos -- a long-waited open problem in computer vision.

See more details in our publication pages.

Prerequisite: Take a machine learning or computer vision class (online class should be okay as well). If you don't take the class and don't have experience but are full of passion in vision and learning, please apply. We welcome you and value your interest!

What you will gain: High quality research experience with ample discussions and GPU resources. International connection for your connecting internships, full time job and graduate adminssions.

Contact in advance: Please email the faculty with your CV and transcript as soon as possible. If you have any previous research experience, please also share the portfolio with us.

Prerequisite: Take a machine learning or a computer vision class (online class should be okay as well. If you're an SNU student, my class will be preferred).

The best way to join us: The best way to join us is to do an internship with us during your undergraduate days.
    If you haven't worked with us (e.g., as an intern), that's okay. We may ask you to go through some initial screening to gauge your experience.

Contact in advance: Please email the faculty with your CV and transcript as soon as possible. If you have any previous research experience, please also share the portfolio with us.

Application schedule: Check out school's admission page for it.