Skip to content
View chenyujiedev's full-sized avatar

Highlights

  • Pro

Block or report chenyujiedev

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
chenyujiedev/README.md
Chen Yujie Computer Vision | Low-Light Vision | Object Detection

Homepage LinkedIn Email


About

I work on low-light computer vision — how detectors and enhancers fail when photons are scarce, and what actually recovers the signal instead of just making the picture brighter.

  • PhD student, Computer Science, Youngsan University (from Sep 2026) · advisor: Prof. Jung Minpo
  • MSc, Computer Science, Youngsan University — GPA 4.4 / 4.5
  • BEng, Computer Science & Technology, Concord University College, Fujian Normal University
  • Previously 4 years as a Senior Software Instructor — full-semester CS courses at ~11 universities and enterprise programs, 1,000+ learners
  • Based in South Korea / China

Research & Publications

DOLENet — Detection-Oriented Low-light Enhancement first author · journal paper in preparation · 2026–present

Enhancement is usually optimized to look good to humans, then bolted onto a detector. DOLENet asks the opposite question: what should enhancement do if its only consumer is a detector? Current work covers multi-seed YOLOv8 / ExDark experiments, gradient-conflict analysis, and TP/FP/FN error decomposition.

MSc Thesis — Deep-Learning Low-light Image Enhancement 2025–2026

Zero-reference training, local illumination balancing and shadow recovery — without adding inference cost.


Featured Projects

paper-Experiment — code behind the thesis work

  • Two loss functions of my own design on the Zero-DCE framework: DarkRegionExposureLoss (region-specific exposure supervision) and HighlightPreservationLoss (guards against over-enhancement)
  • Benchmarked against Retinexformer and RUAS
  • Evaluated on no-reference (NIQE, BRISQUE, LOE) and full-reference (PSNR, SSIM, LPIPS) metrics

chenyujiedev.github.io — personal site: research notes and write-ups


Teaching

Senior Software Instructor, Fujian Guoke Information Technology (2019–2023) — syllabus design, lectures, labs, assignments, exams and student project supervision. Also built a big-data training program for State Grid Corporation of China. Certified Huawei Instructor (HCIA / HCIP).

Area Courses
Programming & SE Java · OOP · Spring Boot · Web Backend · Databases · Microservices · Git
AI & Intelligent Systems Artificial Intelligence · Machine Learning · NLP · PyTorch · Computer Vision · Image Processing
Computer Systems Algorithms · Operating Systems · Computer Architecture · Distributed Systems
Data & Emerging Tech Big Data · Spark · Data Processing · Blockchain

Tech Stack

Python PyTorch OpenCV YOLO NumPy CUDA
Java Spring Boot Vue.js Apache Spark Git Linux


GitHub metrics: repositories, top languages and contribution calendar

Coursework

Graduate coursework kept public for reference: Adaptive CLAHE + Gamma · Advanced Algorithm Design & Analysis · Advanced Distributed Systems · Machine Learning


Snake eating my contribution graph

This repository also hosts the Hexo source for chenyujiedev.github.io.

Pinned Loading

  1. Adaptive-CLAHE-Gamma Adaptive-CLAHE-Gamma Public

    Adaptive CLAHE + gamma correction for illumination correction (graduate coursework).

    Python

  2. chenyujiedev.github.io chenyujiedev.github.io Public

    Personal site: research notes and write-ups. Generated by Hexo.

    Python

  3. paper-Experiment paper-Experiment Public

    Low-light image enhancement: two custom exposure-aware losses on the Zero-DCE framework, benchmarked against Retinexformer and RUAS.

    Python