Thanks for dropping by! I'm based in Korea and build computer vision and machine learning tools, and I'm especially interested in how much you can trust the measurements taken from pose estimation models. I also like making games.

한국에서 컴퓨터 비전과 머신러닝 도구를 만듭니다. 특히 포즈 추정 모델의 측정값을 얼마나 믿을 수 있는지에 관심이 많고, 게임 만들기도 좋아합니다.

poseaudit · Python
Measures how far angles, tilts and lengths derived from pose keypoints disagree with the labels, not just whether the keypoints land near them. With YOLO11n-pose on 322 elbows from 200 COCO val2017 images, 58% of angles are off by 15° or more when the arm segments average under 30 px (37% at 30 to 60 px, 26% above 60 px).
PyPI →

Epokio · Python
Watches ML training runs on your machine or a remote GPU server from the macOS menu bar, or from a browser on Windows, Linux or your phone. No code changes and no account: point it at the folder your runs already write to.
PyPI →

(Also a small browser game: Stella Ball, Play on itch.io.)

• roboflow/supervision#2655: from_yolo parsed whole YOLO pose label rows as polygons, giving wrong boxes. It now reads the box and skips the keypoints. Released in 0.30.8.
• sktime/sktime#11233: Re-enabled tests for two estimators that were skipped by mistake, a follow-up the maintainer requested while reviewing my other (still open) PR.
• MODSetter/SurfSense#2027: Deleting the last connection to a host now also revokes its network access. Hosts still used by other connections keep it.

See all merged pull requests →

If you're working on pose estimation, building games, or just want to talk about either, I'd love to hear from you. Questions and feedback are always welcome.

Contact: bkbkjang@gmail.com