M.S. student at National Taiwan University (GINM) working at the intersection of machine learning, multimedia systems, and AI-generated content. Right now I'm focused on machine unlearning and test-time preference alignment with Prof. Hsuan-Tien Lin. Before that: computer vision, multimedia & HCI at IGVI Lab, and edge-AI engineering at Innovedus.

• 🎓 M.S. Computer Science, NTU GINM (2026–2028) · B.B.A. Technology Management + Interdisciplinary CS & Quant-Finance, NTHU (2021–2026)
• 🔬 Research: machine unlearning · preference alignment · multimedia & creative AI · edge-AI inference
• 🏆 1st Prize — Agent for Truth AI Anti-Fraud Hackathon (2026)
• 📫 masonhuang0320@gmail.com · 🌐 masonhuang.com

• ML & AI safety — machine unlearning, test-time preference alignment, persona alignment, representation learning
• Multimedia & creative AI — AI-generated content, real-time affective visualization, VR / 3D creative environments
• Edge-AI systems — model compression, ONNX deployment, resource-aware inference on NPUs

• PerformaVis: Real-Time Affective Music Visualization Driven by a Pianist's Bodily Expressions  
• 結合肢體與音訊特徵之個人化鋼琴演奏視覺系統設計  
• ImmerseSketch: Transforming Creative Prompts into Vivid 3D Environments in VR  

🛡️ BitoGuard   AML/KYC fraud detection for crypto exchanges — graph-based features + LightGBM with focal loss.	🎹 PerformaVis   Real-time fusion of a pianist's bodily & audio features into affective stage visuals.
🕶️ ImmerseSketch   VR tool that turns free-form creative prompts into immersive 3D environments.	🧟 XMEN Recurrent Dueling DQN (CNN + LSTM) agent for Minecraft zombie combat in Project Malmo.
🌐 Gossip-ES Decentralized evolution-strategies framework — nodes gossip parameter vectors over a graph.	🅿️ Autonomous Parking Planner Benchmarked MPC vs. Hybrid A* vs. SAC deep-RL for automated parking scenarios.

• AI Software Engineer Intern · Innovedus Inc. (2024–2025) — architected KNEO X & Cluster, scalable edge-AI acceleration platforms; built face & voice recognition pipelines for the Kneron NPU.
• Undergraduate Researcher · IGVI Lab, NTHU (2023–2026) — computer vision, multimedia systems, and human–computer interaction.

• TA · Computer Vision Practice with Deep Learning — TAICA CV Program (Fall 2025)
• TA · Introduction to Programming — TAICA Applied AI Exploration Program (Fall 2025)
• Head TA · Software Studio — NTHU (Spring 2025)
• Workshop Instructor · AI Workshop — NTHU Leadership Development Program (Dec 2024)

Last updated on:
July 01, 2026, 04:07 UTC