Yujie Chen

I work on computer vision under low light — the regime where photons are scarce, sensor noise dominates, and the assumptions ordinary detectors were trained on quietly stop holding.

Low-light enhancement is normally optimised to look good to a human, then bolted onto a detector that had no say in what “good” meant. My current work asks the question that split skips: what should enhancement do when its only consumer is a detector? That is DOLENet, my first-author manuscript in preparation. My MSc thesis approached the same territory from the enhancement side, with two exposure-aware losses of my own design.

Research interests — low-light image enhancement · object detection under adverse illumination · zero-reference supervision · no-reference image quality assessment · robustness under distribution shift

Education

• PhD, Computer Science — Youngsan University · 2026–
• MSc, Computer Science — Youngsan University · GPA 4.4 / 4.5
• BEng, Computer Science & Technology — Fujian Normal University · 2015–2019

Previously

Four years as a Senior Software Instructor, independently designing and delivering semester-long computer science courses at roughly eleven universities and corporate programmes, to more than a thousand learners. More →

Email is the most reliable way to reach me. Full details are in my CV.