cv

Bachelor of Science (B.S.)

National Central University, Taiwan

Electrical Engineering & Computer Science (EECS)

GPA / Average Score : 4.1/4.3; 3.9/4; 91/100

DiffQRCoder: Diffusion-based Aesthetic QR Code Generation with Scanning Robustness Guided Iterative Refinement

WACV 2025

Pixel Is Not A Barrier: An Effective Evasion Attack for Pixel-Domain Diffusion Models

Arxiv

Research Assistant

Research Center for Information Technology Innovation, Academia Sinica

Design a novel algorithm for safeguarding images from being unauthorized editing and micmicking via Diffusion Models; Design a novel algorithms for generating aesthetic QR code via training-free diffusion-based method.

• Complete two research projects as co-first author and co-second author respectively.

Research Intern

Institute of Information Science, Academia Sinica

Study adversarial attack and defense methods for computer vision models; Study principles of generative models.

• Study conditional diffusion model behaviors under different conditions.
• Design and implement image intellectual property protection algorithms against malicious usage of Diffusion Models.
• Establish research workflow integrating with Weight&Bias and Notion project management for our research teams.

Network Administrator and Web Developer

EECS Department Office, National Central University

Responsible for department website maintainance and web app developement.

• Develop an automating graduation eligibility reviewer for the EECS Department Office via Chrome Extension and React.js.
• Maintain legacy department website.
• Build a new department website using WordPress and Linux VMs

Fullstack Web Developer

Development and IT Team, Freshman Website Development Program, Office of Student Affairs, National Central University

Build an orientation website for freshmen coming to National Central University (3000+ Users)

• Develop NCUTUBE webpage with interative video playlists, timeline system and voting system, using Vue.js, Nuxt.js and MongoDB.

Second Place in TSMC Group, 11th Hsinchu X Meichu Hackathon

National Tsing Hua University (NTHU) & National Yang Ming Chiao Tung University (NYCU)

Build an AI Cloud Native solution for intelligent X-ray security checking, semiconductor machine maintainance forecasting and employee attendance monitoring systems. Train a YOLO-based detector for prohibited items, achieving rank 1st performance in the leaderboard; Develop a trustworthy interactive natural-language querying system for employee attendance monitoring and automating monthly report generation, using LLMs, PandasAI, and Prompt-engineering; Train timeseries forecasting models to predict machine maintainance date;.

The Dr. Chu Shun‑yi Zyxel Scholarship (Merit‑Based)

National Central University

The most prestigious academic honor for undergraduates at National Central University. Among 3 awardees in the College of EECS.

Excellent Students Scholarship (Merit‑Based)

National Central University

The most prestigious academic honor of the department. Among 2 awardees in the Interdisciplinary Program of Electrical Engineering & Computer Science.

Excellent Club of the Year

National Central University

Awarded 2nd place in the academic category during annual evaluations as club director.

Core Team Member, Instructor

Google Developer Student Club, National Central University

Core team member for Google Developer Group student branch at National Central University, lecturing and advocating Machine Learning topics.

• Host workshop of Introduction to Natural Language Processing on Google Cloud Platform
• Host workshop of Introduction to VertexAI of Google Cloud Platform
• Advocate Machine Learning Study Jams hosted by Google Developer Group

Brain MRI Tumor Segmentation via Segment Anything Foundation Model

Adapting and fine‑tuning Segment Anything Model(SAM) to BraTS brain tumor segmentation task. Reached average 0.72 DICE Score with 12% improvement compared to the original model even if fine‑tuning on a small dataset (3000 image‑mask pairs).

• Term project of BrainHack School 2023, University of Montreal and co-hosts by NCU (Grade: A+, 100/100)
• Segment Anything Model
• Foundation Model
• MRI
• BraTS

RBFNN Self‑driving Car Simulation

A self‑driving car simulation in 2D field using Radial Basis Function Neural Network (RBFNN). Implemented Gradient Descent and Pseudo Inverse to solve the optimization problem.

• Term project of *Neural Networks* course (Grade: A+, 90/100)
• Implemented using plain Python and Numpy

Hopfield Network Implementation

A Hopfield network that could restore states corrupted by random noise. Implemented and investigated the relation between input sample quantity and network capacity, supporting the speculations via theories proposed by Geoffrey Hinton.

• Term project of *Neural Networks* course (Grade: A+, 90/100)
• Implemented using plain Python and Numpy

Wild Cat Breeds Classifier

Recognize common cat breeds in Taiwan, especially being able to identify endangered leopard cats. Reached 0.88 F1‑Score on average.

• Term project of *Introduction to Computer Vision and its Application* course (Grade: A+, 99/100).
• Tensorflow
• Transfer Learning
• EfficientNetV2

Camera Depth of Field Simulator

Calculate Depth of the Field(DoF) and simulate lens bokeh. Using a pre‑trained model to separate out the background, then generate simulated bokeh of background using Gaussian blur parameterized by calculated DoF.

• Term project of *Introduction to Computer Science II* (Grade: A+, 97/100)
• OpenCV
• Java
• MODNet pre‑trained model

Director

Photography Club, National Central University

Lead organizer of the Photography club at NCU

• Give lectures on technical knowledges of cameras and imaging system
• Curate a photography exhibition on campus
• Host a workshop with international photography contest awardee
• Invited as the student representative jury of the school photography competition
• Interviewed by university journalists and published in the university journal.