Yu-Hsuan

About

Hi! I'm Yu-Hsuan Li, an undergraduate student majoring in Computer Science at National Yang Ming Chiao Tung University (NYCU).

Currently, I am a research assistant at the Computational Photography Lab, proudly advised by Prof. Yu-Lun Liu.

My research interests lie in Computer Vision (CV) and Video Diffusion Models (VDM). I am highly motivated to explore video generation and its intersection with world models.

Experience

Internship
June 2026 – Present
Google
Incoming Software Engineering Intern
Google Banqiao Office
Incoming SWE intern at Google Taiwan.
Internship
Sep 2025 – June 2026
TSMC
Software Engineering Intern
Taiwan Semiconductor Manufacturing Company (TSMC)
Built full-stack internal tools and data pipelines to optimize manufacturing and monitoring workflows:
  • Designed and implemented full-stack web applications utilizing React and AG Grid for robust data visualization.
  • Developed efficient backend APIs using FastAPI and optimized data processing pipelines to handle large-scale data.
  • Automated web scraping and data ingestion workflows using scheduled cron jobs for reliable database storage.
Education
Sep 2023 – Jun 2027
NYCU
B.S. in Computer Science
National Yang Ming Chiao Tung University (NYCU)
GPA: 4.28 / 4.3
Rank: 2 / 197 (1.02%)
Advisor: Prof. Yu-Lun Liu

News

Publications

YoCausal teaser
arXiv 2026

YoCausal: How Far is Video Generation from World Model? A Causality Perspective

You-Zhe Xie*, Yu-Hsuan Li*, Jie-Ying Lee, Kaipeng Zhang, Yu-Lun Liu†, Zhixiang Wang†

*Equal contribution  †Corresponding authors

The first benchmark evaluating causal understanding in video generation models. Our benchmark can incorporate any real-world video at zero cost, making it arbitrarily extensible to easily assess video generation models' understanding of diverse types of causality.

Projects

Guitar Tutor for the Blind

Guitar Tutor for the Blind

Sep 2025
Python FastAPI React Google AI
  • 2025 MC Hackathon × Google — AI-powered mobile web app for visually impaired users to learn guitar
  • Real-time audio feedback and voice-guided instruction via Google AI APIs
  • Python / FastAPI backend + React frontend over local HTTPS to Android
NYCU Deep Learning Labs

NYCU Deep Learning Labs

Feb 2025
Python PyTorch Diffusion RL
  • 7 labs for NYCU Deep Learning & Practice course — scored 97/100 and 96/100
  • Semantic segmentation (U-Net), image inpainting (MaskGIT), video prediction (VAE)
  • Conditional image generation (DDPM) and reinforcement learning (DQN, PPO, A2C)
LLM-Assisted Paper Search

LLM-Assisted Paper Search

Jun 2024
Python LangChain RAG Discord
  • Discord bot for natural language arXiv paper retrieval and bilingual summarization
  • RAG pipeline with LangChain + GPT-3.5; paper database built via Selenium scraping
  • 96.6% content relevance across 30 test cases
2D Zombie Survival Game

2D Zombie Survival

Aug 2023
Unity C# Game Dev 2D Physics
  • Solo-developed 2D survival shooter — survive 15 minutes of escalating zombie waves
  • Collect XP to level up; pick 1-of-3 upgrades from 11 skills and 5-tier weapons
  • Built entirely from scratch in C#: status effects (freeze / slow / burn / knockback), coroutines, and late-game frame-rate optimization
Play

Teaching

Spring 2026

Computer Organization (CS10014)

Teaching Assistant · National Yang Ming Chiao Tung University

Responsible for designing and grading Labs 3 – 5:

  • Lab 3 Single-Cycle CPU — implement a complete RISC-V single-cycle processor in Verilog supporting 20 instructions (arithmetic, memory, branch, jump)
  • Lab 4 Pipelined CPU — extend Lab 3 into a 5-stage pipeline with a Hazard Detection Unit (stall on load-use) and a Forwarding Unit (resolve data hazards)
  • Lab 5 Cache Manager — implement a 2-way set-associative cache (1 KB, 16-byte blocks, LRU, Write Allocate) in C++ with correct miss-count tracking

Contact

Feel free to reach out — I'm always happy to chat about research and collaboration.

I'm open to research collaborations, internship opportunities, and interesting conversations.