Home
取消

Profile Picture

Yi-Hsiang (Sean) Ho | 何義翔

I'm a Computer Science student passionate about AI/ML, Systems Engineering, and Software Development. I recently interned at Appier as a Backend Engineer and I have extensive experience in full-stack development, Linux system administration, and computer vision research.

Experience & Education

NTU
M.S. in Computer Science and Information Engineering Sep 2026 - Present

National Taiwan University (NTU)

Appier
Backend Engineer Intern Jul 2026 - Present

Appier

Trend Micro
AI Engineer Intern Jul 2025 - Aug 2025

Trend Micro

NYCU
Research Student Jul 2024 - Present

Computational Photography Laboratory, NYCU

NYCU
B.S. in Computer Science Jul 2022 - Jun 2026

National Yang Ming Chiao Tung University (NYCU)

Projects

CloudNest
CloudNest
3rd Place in Hacker Group (CloudMosa), 2025 Meichu Hackathon
FastAPI, PostgreSQL, Docker

A labor matching platform for underserved regions. Architected and implemented a RESTful API backend and database to power a Cloud Phone-based platform.

Bare-metal Kernel on Raspberry Pi
Bare-metal Kernel on Raspberry Pi
Operating System Capstone
C, Assembly

Developed a 64-bit multitasking OS kernel from scratch. Engineered memory management, UART drivers, and a multi-tasking scheduler.

fftconvolve2d
fftconvolve2d
Numerical Software Development
C++, Python, pybind11

Engineered a cross-language faster 2D convolution solution using Fast Fourier Transform.

Meichu Hackathon Web
Meichu Hackathon Web
Vue.js, Tailwind CSS

Co-led development for the 2024 event platform. Facilitated sign-ups for 300+ participants.

Exposure Correction
Exposure Correction
Image Processing
Python, OpenCV, PyTorch

Optimized Zero-DCE, a low-light enhancement approach, to address both over- and under-exposure challenges. Achieved a 144% improvement in PSNR and 54.5% in SSIM.

Deep Learning for Visual Recognition
Deep Learning for Visual Recognition
Selected Topics in Visual Recognition using Deep Learning
Python, PyTorch

Implemented computer vision models (ResNet, Faster R-CNN, Mask R-CNN, PromptIR) for classification, segmentation, and restoration. Optimized performance using data augmentation and cross-validation.

Deep Learning Labs
Deep Learning Labs
Deep Learning
Python, PyTorch, NumPy

Implemented seven labs covering backpropagation in NumPy, UNet segmentation, MaskGIT inpainting, conditional VAE video prediction, DQN variants, conditional diffusion, and A2C/PPO. Reached 0.934 Dice, 29.1 FID, and a Pong score of 19.95.

Meichu Game Web
Meichu Game Web
Vue.js, Cloudflare

Official web portal for the annual Meichu Game tournament. Re-architected into a monorepo and implemented edge routing via Cloudflare Workers for seamless historical site traffic.

Band System
Band System
Introduction to Database Systems
React, Flask, PostgreSQL, AWS

A full-stack application to pair up compatible band members with dynamic filtering, profile discovery, and invitation workflows. Deployed on AWS.

Text-based Dungeon RPG
Text-based Dungeon RPG
Data Structures and Object-oriented Programming
C++

A CLI game utilizing the concept of OOP. Features include character occupations, battle mechanics, and file saving/loading.