01 · Context
The data regime, the task boundary, and the failure that motivates the method.
About
An archive, not a summary feed.
Reviews here keep the parts of a paper that stay useful once the novelty language fades.
Jisang Lee
M.S. student at AIMLab, Hanbat National University, working across remote sensing, medical imaging, forgery detection, and open-vocabulary vision.
Current interests: segmentation, remote sensing, medical imaging, forgery detection, and open-vocabulary recognition.
- Role
- M.S. Student in Computer Engineering
- Lab
- AIMLab
- Advisor
- Prof. Haneol Jang
- Based in
- Daejeon, Republic of Korea
How a review is written
02 · Mechanism
Inputs, transformations, supervision, outputs. Not a list of module names.
03 · Evidence
What the authors measured, separated from what the review infers.
04 · Limits
Confounds, missing controls, cost, and the next experiment worth running.