CVPR 2026 doScenes Instructed Driving Challenge
NudgeVAD: Language-Nudged End-to-End Driving via FiLM Residuals
Ablation Track: 1st Place Winner; Full Multimodal End-to-End Track: Runner-up.
Yu-Hsiang Chen
I am a master's student in Computer Science at National Yang Ming Chiao Tung University, advised by Prof. Yi-Ting Chen in HCIS Lab. My research focuses on autonomous driving safety, heterogeneous traffic simulation, computer vision, deep learning, and robotics.
I was a visiting researcher at the MSC Lab, University of California, Berkeley, working with Prof. Masayoshi Tomizuka and Prof. Yi-Ting Chen on data-efficient and controllable traffic simulation. Previously, I graduated from National Tsing Hua University in the top 1% of my class.
I am open to research collaboration and seeking Fall 2027 PhD opportunities.
News
- HetroD accepted to ICRA 2026.
- Organized ECCV and ICCV DriveX Workshops and initiated the Heterogeneous Traffic Simulation Challenge.
Publications
ICRA 2026
HetroD: A High-Fidelity Drone Dataset and Benchmark for Heterogeneous Traffic in Autonomous Driving
Under Review, ECCV 2026
ECoSim: Data-Efficient Fine-Tuning for Controllable Traffic Simulation
ICCV 2025 DriveX
V2X-based Logical Scenario Understanding with Vision-Language Models
NSTC 2023
Low-Risk Arbitrage Strategy via Inter-Cryptocurrency Correlation Analysis in High-Frequency Trading
Professional Experience
Incoming Video Deep Learning Research Intern, Synology
06/2026 - 12/2026
MSC Lab, University of California, Berkeley
06/2025 - 12/2025
- Pioneered a heterogeneous dataset and simulation benchmarks for autonomous driving simulation, accepted to ICRA 2026.
- Engineered a data-efficient, controllable traffic simulation framework using diffusion and autoregressive models for realistic closed-loop performance.
HCIS Lab, National Yang Ming Chiao Tung University
09/2024 - 06/2026
- Co-developed safety validation systems for autonomous driving, integrating V2X-based logical scenario retrieval via VLMs and VLM-guided end-to-end driving.
- Teaching assistant for Introduction to Artificial Intelligence and Physical AI.
R&D Software Engineer, Geniideas
06/2023 - 04/2026
- Spearheaded an automated piping ISO system, reducing manual overhead and improving installation workflows for enterprise clients including TSMC and Micron.
- Architected computer vision solutions for laser welding anomaly detection and autonomous deep learning classification for crystal oscillator operations.
IS Lab & Institute of Service Science, National Tsing Hua University
09/2022 - 06/2024
- Designed a high-frequency trading framework to capture sudden market fluctuations; funded by NSTC and published as an undergraduate research project.
- Developed data-crawling APIs for 400,000+ luxury goods entries and built React web applications for campus sustainability and tour workflows.
Education
09/2024 - 06/2026
National Yang Ming Chiao Tung University
Master of Computer Science, HCIS Lab. Advisor: Prof. Yi-Ting Chen.
09/2020 - 06/2024
National Tsing Hua University
Bachelor of Technology Management, IS Lab. Advisor: Prof. Hung-Min Sun. Ranked 2/65, top 1% of class.
Awards & Service
- 5x Presidential Award, National Tsing Hua University, 2020-2024
- Outstanding Graduate Award, National Tsing Hua University, 2024
- Scholarship for Excellent Students to Overseas Internship, Ministry of Education, Taiwan, 2024
- College Student Research Scholarship, National Science and Technology Council, 2023
- Organizer, ECCV & ICCV DriveX Workshops
- Initiator, Heterogeneous Traffic Simulation Challenge
- Reviewer: NeurIPS, CVPR, ECCV, ICRA, IEEE RA-L, T-Rel, IJPRAI
- Skills: Python, C++, C, JavaScript, PyTorch, Qt, Git, LaTeX, Linux, MultiCharts
Contact
I am open to research collaboration and seeking Fall 2027 PhD opportunities.