‘Science plus arts multiplied by faith then divided by the number of our ethnic race.’ — Nujabes, feat. Shing02, Luv(sic) pt2

About Myself

Hi! I'm a Master's student at National Taiwan University 🇹🇼, where I am a member of the Communications and Multimedia Lab and advised by Prof. Robin Bing-Yu Chen. I am an AI researcher with international experience in sim-real learning and computational solvers for high-performance, physical and embodied systems. My research spans 1. graphics: simulation and animation production; 2. robotics: manipulation, dynamics, and sensing; and 3. audio: signal processing and room acoustics, focusing on physics-informed ML, neural PDE/ODE, reinforcement and robot learning, graph, and generative/multimodal models.

Research Areas

I like systematically working through critical problems that sit between disciplines, and my current interests span three core areas:

  • Visual Computing & Physics Simulation — developing neural PDE solvers (e.g., Walk-on-Spheres Monte Carlo for complex geometry) and differentiable physical simulation frameworks (hyper-elastic projective dynamics with positive-definite Hessian projection), with applications spanning CAD-driven robotic fabrication and animation workflows such as editable 3D Gaussian Splatting.

  • Embodied Systems — training Sim2Real reinforcement learning policies in Isaac Gym/Lab for robotic manipulation on Kuka, UR10, and Hiwin arms, and building large-scale digital twins that fuse multimodal sensing (WiFi, point cloud, RGBD) with graph federated learning foundation models for privacy-preserving human–scene interaction on supercomputers like Fugaku.

  • Audio Signal Processing — combining classical DSP with learned latent-space representations to design audio watermarks robust to neural resynthesis, and modeling spatial audio through room impulse responses (RIRs) in complex room acoustics.

Internships 🇫🇮 🇯🇵 🇹🇼

Along with my work at NTU, I've had the chance to do research internships and build collaborations across several international groups. I spent time at Aalto University through the 2026 Aalto Science Institute (AScI) Summer Research Programme (QuML Group, Espoo, Finland 🇫🇮; advised by Prof. Vikas Garg), and at RIKEN's Center for Computational Science (R-CCS) through the 2025 R-CCS International HPC Computational Science Internship Program (Large-Scale Digital Twin Research Team, Kobe/Osaka, Japan 🇯🇵; under Team Principal Prof. Hirozumi Yamaguchi and Prof. Hamada Rizk of Osaka University). On the industry side, I've worked in R&D at MoonShine Animation Studio (with Dir. Shaune Jan) and Delta Electronics.

Collaborations 🇺🇸 🇨🇦 🇫🇮 🇸🇬 🇹🇼

I have co-worked with cross-cultural (20+ nationalities) researchers from the U.S. DoE, ETH Zurich, NUS, UBC, UCLA, UCL, TU Delft, UOsaka, Aalto, etc. My ongoing academic collaborations include Prof. Peter Yichen Chen (UBC, Canada 🇨🇦), Dr. Tiantian Liu (Meshy AI, US 🇺🇸), Dr. Ziqiu Zeng (NUS, Singapore 🇸🇬), Prof. Fan Shi (NUS, Singapore 🇸🇬), and Prof. Yu-Lun “Alex” Liu (NYCU, Taiwan 🇹🇼), as listed in my CV. My broader ongoing collaborations also include Prof. Yun-Nung Chen (NTU, Taiwan 🇹🇼), Prof. Hung-Yi Lee (NTU, Taiwan 🇹🇼), Prof. Shang-Tse Chen (NTU, Taiwan 🇹🇼), Kyung Lee (Aalto, Finland 🇫🇮), Nils Meyer-Kahlen (Aalto, Finland 🇫🇮), Eloi Moliner (Aalto, Finland 🇫🇮), and Prof. Xiang “Anthony” Chen (UCLA, US 🇺🇸). Working on state-of-the-art research with international teams has been one of the most rewarding parts of being in research, and is something I now actively look for.

Before Master

I also worked closely with multiple academic groups during my undergraduate studies and co-authored 8 conference papers in top-tier AI/CS venues. These include the Urban Science & Computing Lab led by Prof. Hsun-Ping Hsieh, and RAC-Coon (Robot Aided Creation and Construction Research Group) together with Prof. Yang-Ting Shen, Prof. Chia-Ching Yen, and Prof. Wen-Yu Su.

Community Services/Networks

I contribute to the research community as an ACM reviewer for SIGGRAPH Asia, CHI, HRI, CSCW, and The Web Conference (WWW). Separately, I serve as an IEEE reviewer for ITSC.

If you are looking to connect with excellent researchers, I would also be happy to recommend the following friends:

  • Nikoleta E. Glynatsi — RIKEN Center for Computational Science (R-CCS), Japan
    Game Theory and Evolutionary Dynamics
  • Nassim Beladel — ETH Zurich, Switzerland
    Self-Supervised Learning and Generative Social Choice in Politics and Communities
  • Louka Verstraete — ETH Zurich, Switzerland
    FPGA Systems and Quantum Physics, with a focus on Fault Tolerance and Quantum Error Correction
  • Antoine Kempf — CEA, France
    High-Performance Computing and Numerical Simulation, including Neutron Transport, Kokkos, and Computational Fluid Dynamics
  • Pei-Chi Tsai University College London (UCL), UK
    Cities, Sustainable Strategy, Urban Heat Island, Data Analysis, and Extended Reality
  • Yen-Shan (Lily) Chen — National Taiwan University (NTU), Taiwan
    AI Safety and Large Language Models, with a focus on Trustworthy AI, LLM Robustness, Agentic Safety, and Mechanistic Interpretability
  • Tzu-Hsin Hsieh — Delft University of Technology (TU Delft), the Netherlands
    Augmented and Virtual Reality, Human–Computer Interaction, and Artificial Intelligence
Music — YouTube Playlist

Update sometimes, depends on your donation or my mood.

Collaboration is one of the parts of research I enjoy most — technical discussions, sharing expertise, and figuring out what we can build together. If any of the above resonates, let's discuss!

Outside of research: graffiti (in legal areas), live houses/music festivals, and traveling.

Publications

Domain #AI #Robotics #Graphics #HPC #Audio
Method #Sim-Real Transfer #Graph #RL #DSP #Physics-informed ML

Selected (6/14)

DiffPhD teaser

DiffPhD: A Unified Differentiable Solver for Projective Heterogeneous Materials in Elastodynamics with Contact-Rich GPU-Acceleration
ACM Transactions on Graphics (TOG), SIGGRAPH ASIA 2026 (1st Author, Q1 Journal: top-tier in graphics, IF=13).
preprint arXiv:2605.14526 PDF Cite
Shih-Yu Lai, Sung-Han Tien, Jui-I Huang, Yen-Chen Tseng, Yi-Ting Chiu, Siyuan Luo, Ziqiu Zeng, Fan Shi, Peter Yichen Chen, Tiantian Liu, Yu-Lun Liu, Bing-Yu Chen
#Graphics #Robotics #Sim-Real Transfer


ITSC 2026 teaser

MORPH-U: Multi-Objective Resilient Motion Planning for V2X-Enabled Autonomous Driving in High-Uncertainty Environments via Simulation
IEEE ITSC 2026 (1st Author, Regular Paper). PDF
Shih-Yu Lai
#AI #Robotics

DASTGCN-SNN-DDPG teaser

Architectural CAD/CAM in Robotic Fabrication via Differentiable Energy Optimization and Few-Shot Spatiotemporal Graph RL Policy
ACM SIGGRAPH Asia 2025 (1st Author, Technical Communications). PDF Cite
Shih-Yu Lai, Chia-Ching Yen, Yang-Ting Shen, Bing-Yu Chen
#AI #Robotics #Graphics #Graph #RL #Physics-informed ML

GAT-DQN teaser

Graph Q-Learning for Automatic Assembly in Design-to-Construction
ACM SIGGRAPH Asia 2025 (1st Author, Poster). PDF Cite
Shih-Yu Lai, Chia-Ching Yen, Yang-Ting Shen, Bing-Yu Chen
#AI #Robotics #Graphics #Graph #RL

EUPHORIA teaser

EUPHORIA: Efficient Universal Planning via Hybrid Optimization for Robust Industrial Robotic Assembly
ArXiv 2026 (1st Author, Preprint). PDF
Shih-Yu Lai, Chia-Ching Yen, Yang-Ting Shen, Peter Yichen Chen, Yu-Lun Liu, Bing-Yu Chen
#AI #Robotics #Physics-informed ML

UMEDA teaser

UMEDA: Unified Multi-modal Efficient Data Fusion for Privacy-Preserving Graph Federated Learning via Spectral-Gated Attention and Diffusion-Based Operator Alignment
ArXiv 2026 (1st Author, Preprint). PDF
Shih-Yu Lai, Hirozumi Yamaguchi, Shang-Tse Chen, Yu-Lun Liu, Bing-Yu Chen
#AI #HPC #Graph #Physics-informed ML

Others (8/14)

ISC 2026 teaser

Accelerating Large-Scale Federated Graph Foundation Model Pre-training for Secure Sensing via Hierarchical Parallelism
ISC High Performance 2026 (1st Author, Research Poster, Presented). PDF
Shih-Yu Lai, Fukuharu Tanaka, Hamada Rizk, Hirozumi Yamaguchi, Bing-Yu Chen
#AI #HPC #Graph

LATENT-MARK teaser

LATENT-MARK: An Audio Watermark Robust to Neural Codec Compression
Interspeech 2026 (Co-1st Author, Long Paper; Scores: 5, 5, 5, 4). PDF Code
Yen-Shan Chen, Shih-Yu Lai, Ying-Jung Tsou, Yi-Cheng Lin, Bing-Yu Chen, Yun-Nung Chen, Hung-Yi Lee, Shang-Tse Chen
#AI #Audio #DSP

M3 teaser

M3: Recommendation via Attention-Graph Cluster Q-Learning with Multi-Scale Spatial Heterogeneity for Multi-Purpose, Multi-Stakeholder Green Attractions in Transportation
ACM SIGSPATIAL 2025 (1st Author, Research Full Paper). PDF Cite
Shih-Yu Lai, Tzu-Hsin Hsieh, Pei-Chi Tsai, Chao-Chun Kung, Sing-Kai Ling, Hsun-Ping Hsieh
#AI #Graph #RL

LINKin-PARK teaser

LINKin-PARK: Land Valuation Information and Knowledge in Predictive Analysis and Reporting Kit via Dual Attention-DCCNN
ACM CIKM 2024 (2nd Author, Demo Paper). PDF Cite
Teng-Yuan Tsou, Shih-Yu Lai, Hsuan-Ching Chen, Jung-Tsang Yeh, Pei-Xuan Li, Tzu-Chang Lee, Hsun-Ping Hsieh
#AI

GreenSpot teaser

GreenSpot: Improving Public Transport with GIS-Based AR and Cluster-GCN Recommendation
ACM SIGSPATIAL 2024 (1st Author, Demo Paper). PDF Cite
Shih-Yu Lai, Tzu-Hsin Hsieh, Sing-Kai Ling, Pei-Chi Tsai, Chao-Chun Kung, Hsun-Ping Hsieh
#AI #Graph

ACCEPT teaser

ACCEPT: A Context-Sensitive, Configurable, and Extensible Prediction Tool using Grid-based Data Processing and Neural Networks in Geospatial Decision Support
ACM SIGSPATIAL 2024 (2nd Author, Demo Paper). PDF Cite
Teng-Yuan Tsou, Shih-Yu Lai, Jung-Tsang Yeh, Hsuan-Ching Chen, Pei-Xuan Li, Tzu-Chang Lee, Hsun-Ping Hsieh
#AI

Footprints of Travel teaser

Footprints of Travel: AIoT and AR Enhanced Tourist Gaming Experience in Unmanned Cultural Sites
ACM CHI 2024 - EA (1st Author, Student Game Competition). PDF Cite
Shih-Yu Lai, Dai-En Liu, Zong-Fan Yeh, Chun-Che Hsu, Fu-Yao Chang, Ming-Zhi Tsai, Ruei-Hong Lin
#AI

UrbanCrowds teaser

Enhancing Urban Crowd Monitoring through Predictive Modelling System with Diverse Geospatial Datasets
ACIS 2023 (3rd Author, Proceeding). PDF Cite
Hsun-Ping Hsieh, Tzu-Chang Lee, Shih-Yu Lai, Pei-Chi Tsai, Tzu-Hsin Hsieh
#AI
No publications match the selected tags.

Thesis

M.S. Thesis (in progress): Neural Physics Solver in Complex Geometry Processing with Walk-on-Spheres Monte Carlo Method for Partial Differential Equations.

Research & Work Experience

Research Intern, QuML Group (1% acceptance rate; fully funded; applicants from 135 countries)

Jun. 2026 – Sep. 2026
Aalto University Aalto University, School of Science — Helsinki, Finland
  • Researched scalable physics-informed learning, neural PDEs/ODEs, and generative modeling for complex geometries under Prof. Vikas Garg.
  • Selected as the first Taiwanese participant in the program’s five-year history.
  • 2026 Aalto Science Institute International Summer Research Programme | Project: “Machine Learning: Foundations and New Frontiers”.

Visiting Researcher

Jun. 2026 – Sep. 2026
Aalto University Aalto Acoustics Lab, Aalto University, Espoo, Finland
  • Devising supervised flow-matching models that learn room impulse responses for robust sim-to-real transfer and cross-room generalization, in collaboration with Kyung Yun Lee, Dr. Nils Meyer-Kahlen, and Dr. Eloi Moliner.

HPC Research Intern (Supercomputer Fugaku), Large-Scale Digital Twin Research Team (2% acceptance rate; fully funded)

Jul. 2025 – Aug. 2025
RIKEN RIKEN Center for Computational Science — Kobe–Osaka, Japan
  • Researched diffusion neural operators, ODE dynamics, and world models for robust physics-informed sensing under Prof. Hirozumi Yamaguchi.
  • Devised graph federated learning for multimodal sensing on supercomputer Fugaku, scaling to 10k clients with enhancing 15x training throughput.
  • Built edge/cloud distributed computing pipelines with graph federated learning foundation models and differential privacy for human–scene interaction.

AI Research Engineer, R&D Team

Sep. 2025 – Dec. 2026
MoonShine MoonShine Animation Studio — Taipei, Taiwan (Dir. Shaune Jan; Prof. Xiang “Anthony” Chen, UCLA)
  • Developed synthetic data-based world model, VLM and 3D/4DGS-based creator tools for animation, VFX, and entertainment production workflows.
  • Designed multi-agent MLLM-as-a-judge for efficient artist-supervisor review under Dir. Shaune Jan and Prof. Xiang “Anthony” Chen (UCLA), achieving a 74% preference rate among practitioners.

Research Assistant (Interactive Graphics Lab)

Nov. 2024 – Aug. 2027
NTU Communications & Multimedia Lab, National Taiwan University (Advisor: Prof. Robin Bing-Yu Chen)
  • Differentiable Projective Dynamics hyper-elastic mesh solver with positive-definite Hessian projection; 1.65×–2.26× acceleration via stable clamping strategies.

Research Assistant

Apr. 2022 – Jun. 2024
RAC-COON Robot-Aided Creation & Construction Research Group — Synchronization Lab (Prof. Yang-Ting Shen, Prof. Chia-Ching Yen, Prof. Alvin Wen-Yu Su)
  • Robotic reinforcement learning on Kuka/UR10/Hiwin in Omniverse Isaac Gym/Lab; graph RL for geometric brick assembly by mobile manipulators.
  • ROS2 systems integrating SLAM, supervised learning (YOLO, LSTM), and RL for motion control, planning, and obstacle avoidance.

Deep Learning Intern, Power and System Business Group

Jul. 2023 – Aug. 2023
Delta Delta Electronics — Taoyuan, Taiwan
  • Developed a PyTorch3D point-cloud autoencoder using 74k meshes and an XGBoost surrogate model for finite-element-based magnetic-core design, reducing prediction error to 1% of the formula-based baseline.
  • Implemented the PyTorch3D training pipeline with Docker and NVIDIA V100 infrastructure.

Honors & Awards

2026 Foxconn Tech Award

Jun. 2026
  • Selected for the Foxconn Tech Award ($100,000) for my undergraduate first-author publication — combining attention-based graph clustering with reinforcement learning (Dueling DDQN).
  • The honored work is M3 (Research Full Paper), presented at ACM SIGSPATIAL 2025 and applied to personalized green-travel recommendations for public transport.

2023 AIoT Innovation Challenge

Feb. 2023 – Jun. 2023
  • Secured 3rd Place & Sponsors Award ($30,000) from NXP Semiconductors (Awards).
  • Built a microcontroller-based SensorHub (ADI/NXP/Arduino) integrating DHT11/SCD40/PMS5003T via chips communication(I²C/UART); developed calibration tests with real-time logging in HVAC/fan/lighting control linked to Unity digital twin for vision model based crowd regulation.

Overseas Conference Travel Grant

2026
  • Travel support recipient for attending the ISC High Performance conference 2026 (June 22-26, Hamburg, Germany).

2025 Overseas Conference Travel Grant

Dec. 2025
  • Graduate student travel support recipient ($1,000) for SIGGRAPH Asia 2025 conference (December 16 – 18, Hong Kong).

Graduate Student Scholarship

2025 – 2026
  • Recipient of the graduate student scholarship for the 2025-2026 academic year.

Academic & Community Service

Reviewer

Jan. 2024 – Jan. 2025
ACM ACM @ SIGGRAPH ASIA’26, CHI’25, HRI’25, CSCW’24, WWW’25
  • Reviewed 9 full- and short-paper submissions in animation/simulation, human–computer/robot interaction and web retrieval.

Reviewer

Mar. 2026 – Apr. 2026
IEEE IEEE @ ITSC’26
  • Reviewed 1 full-paper submission in intelligent transportation systems.

Lead Technical Specialist

Jan. 2025 – Jun. 2027
NTU AI Club NTU AI Club @ Taipei, Taiwan
  • Spearheaded academic research initiatives and lectures on AI; designed and mentored advanced technical workshops in AI research.

Mandatory Military Service

Jul. 2024 – Nov. 2024
Taiwan Air Force Taiwan Air Force