Yichen Shi
Shanghai, China · shiyichen@sjtu.edu.cn · Homepage · Google Scholar
Summary
PhD student in Electronic Information and Electrical Engineering at Shanghai Jiao Tong University, focusing on AI (LLM) for EDA, multimodal reasoning, and biometric security. Experienced in building datasets, benchmarks, multimodal pipelines, and applied machine learning systems through both academic research and industry internships. Seeking research or engineering roles in AI, multimodal systems, EDA intelligence, or applied machine learning.
Education
- Shanghai Jiao Tong University, Ph.D. in Electronic Information and Electrical Engineering, expected 2027
- Shijiazhuang Tiedao University, M.S. in Computer Science, 2023
- Shijiazhuang Tiedao University, B.S. in Computer Science, 2020
Research Interests
- AI (LLM) for EDA: analog and mixed-signal circuit understanding, multimodal reasoning, circuit datasets and benchmarks, knowledge-enhanced auto-design
- Computer Vision and Security: face anti-spoofing, face forgery detection, explainable multimodal models
Experience
Algorithm Intern, Shukun Technology, Sep 2023 – Jun 2024 Worked on Segment Anything Model (SAM) and multimodal large language models for applied vision tasks.
Algorithm Intern, Ping An Technology, Jun 2023 – Sep 2023 Conducted research on multimodal learning and practical model development.
Algorithm Intern, Hikvision, Feb 2022 – Jun 2023 Worked on face anti-spoofing and face detection algorithms for biometric security scenarios.
Selected Publications
- FaceShield: Explainable Face Anti-Spoofing with Multimodal Large Language Models. AAAI 2026.
- SHIELD: An Evaluation Benchmark for Face Spoofing and Forgery Detection with Multimodal Large Language Models. Visual Intelligence, 2025.
- AMSnet-KG: A Netlist Dataset for LLM-based AMS Circuit Auto-Design Using Knowledge Graph RAG. ACM TODAES, 2025.
- AMSbench: A Comprehensive Benchmark for Evaluating MLLM Capabilities in AMS Circuits. arXiv, 2025.
- AMSnet 2.0: A Large AMS Database with AI Segmentation for Net Detection. arXiv, 2025.
- AMSNet: Netlist Dataset for AMS Circuits. LAD 2024 Best Paper Award.
- Symbol and Footprint Database for Electronic Components by Agentic Recognition and Generation. PRCV 2025.
Projects and Research Highlights
- Built AMSNet, AMSnet 2.0, and AMSnet-KG, a line of datasets and data-centric infrastructure for AMS circuit understanding, netlist generation, and LLM-based circuit auto-design.
- Developed AMSbench, a benchmark for evaluating multimodal large language models on AMS circuit perception, analysis, and design tasks.
- Contributed to FaceShield and SHIELD, covering explainable face anti-spoofing, face forgery detection, multimodal evaluation, and model reasoning.
- Co-founded FEATs (Future EDA and AI Techniques Seminar) to support academic exchange among early-career researchers in AI and EDA.
Technical Skills
- Programming: Python, C++
- Domains: machine learning, multimodal learning, large language models, computer vision, EDA, AMS circuits
- Research Workflow: dataset construction, benchmarking, model evaluation, experiment design, academic writing
Awards
- LAD 2024 Best Paper Award, for AMSNet: Netlist Dataset for AMS Circuits
- Lanqiao Cup, Hebei Province, First Prize (Ranked 1st), 2020
- HBCPC (Hebei Provincial Programming Contest), Gold Medal (Top 10), 2020
- ACM-ICPC Xi’an Regional, Bronze Medal, 2019
- CCF-CCSP North China Regional, Bronze Medal, 2019
- HBCPC (Hebei Provincial Programming Contest), Gold Medal (Top 10), 2019
- ACM-ICPC Ningxia Regional, Bronze Medal, 2018