Hi there! π I am Chang Yu (δΊη ), a Postdoc researcher at the College of Engineering, Westlake University, working with Chair Prof. Stan Z. Li (IEEE Fellow). I received my Ph.D. in 2024 from the National Laboratory of Pattern Recognition (NLPR), Institute of Automation, Chinese Academy of Sciences (CASIA), advised by Prof. Zhen Lei (IEEE Fellow). Before that, I received my B.E. in Automation from Xiβan Jiaotong University.
My research sits at the intersection of AI for Science and World Model. I focus on uncovering intrinsic structural priors in complex systems β hierarchical composition, the biological Central Dogma, and dynamical priors β and using them to constrain deep representation and generation models, bridging visual perception with biological modeling.
If you are interested, feel free to contact me for future collaborations through Email.
π Research Interests
My work centers on discovering intrinsic structural priors in complex systems and using them to guide deep representation and generation. Current directions:
a) AI for Science
- Cross-omics modeling along the Central Dogma (DNA β RNA β protein)
- Single-cell dynamics and context-aware biological representation
- Structure-based drug design and molecular modality expansion
b) Hierarchical Part-Whole Representation
- Capsule networks for unsupervised part discovery
- 3D face modeling beyond 3DMM
- 2Dβ2.5Dβ3D pathway in neural networks (Marr-inspired)
c) Unified Dynamics on Data Manifolds
- ODE discovery, RNA velocity inference, and AI-generated video detection
- Manifold decoupling for cross-task dynamics modeling
π₯ News
- 2026: β¨β¨β¨ Geometric Flow Grounding (first author) accepted at ICML 2026 as Oral, Top 0.7%.
- 2026: β¨β¨β¨ CDBridge (first author) accepted at ICLR 2026.
- 2026: β¨β¨β¨ SYNC (co-corresponding) accepted at ICLR 2026.
- 2026: β¨β¨β¨ SteinsGate (co-corresponding) accepted at ICLR 2026.
- 2026: β¨β¨β¨ VitaMol (co-corresponding) accepted at Pattern Recognition.
- 2025: β¨β¨β¨ DaCapo (co-corresponding) accepted at CVPR 2025.
- 2025: π Started as Postdoc at College of Engineering, Westlake University, working with Prof. Stan Z. Li.
- 2024.06: πππ Obtained Ph.D. degree from CASIA, advised by Prof. Zhen Lei.
- 2023: β¨β¨β¨ Graphics Capsule (first author) accepted at CVPR 2023.
- 2022: β¨β¨β¨ HP-Capsule (first author) accepted at CVPR 2022.
- 2022: β¨β¨β¨ Beyond 3DMM accepted at T-PAMI.
π οΈ Research Projects
Central DogmaβGuided Cross-Omics Learning: From DNA to Cell
Open Project of the State Key Laboratory of Multimodal Artificial Intelligence Systems (MAIS), Institute of Automation, Chinese Academy of Sciences
Principal Investigator Β· 2025 β 2026
Deep Transformation of Protein SequenceβStructure Relationships and Protein-Target Drug Design
Ministry of Science and Technology of China, STI2030 Major Project on New-Generation Artificial Intelligence
Core Member Β· 2025 β 2027 Β· Responsible for mid-term report and multi-team coordination
Prompt Mechanism in Latent Diffusion for Text-to-Image Generation
Research Collaboration with Huawei
Research Intern Β· 2022 β 2023 Β· Improving text-to-image accuracy from both quality and semantic perspectives
π Publications
Chang Yu highlighted. † Equal contribution. * Corresponding author.
Geometric Flow Grounding: A Unified Manifold Decoupling Framework for Dynamics Discovery and Verification (Oral, Top 0.7%)
Chang Yu, Yuxuan Luo, Yixuan Du, et al.
CDBridge: A Cross-omics Post-training Bridge Strategy for Context-aware Biological Modeling
Chang Yu, Siyuan Li, Zicheng Liu, Jingbo Zhou, Xianglong Guo, Kai Yu, Yuqing Zhou, Ken Li, Zelin Zang, Zhen Lei, Stan Z. Li
GraVQ: Fine-Grained Cross-Omics Translation via Group-Advantage Vector Quantization
Chang Yu, Xianglong Guo, Jingjing Jiang, Siyuan Li, Jingbo Zhou, Changxi Chi, Zelin Zang, Zhen Lei, Stan Z. Li
SYNC: Measuring and Advancing Synthesizability in Structure-Based Drug Design
Yunfan Liu, Lirong Wu, Zhifeng Gao, ..., Chang Yu*, Stan Z. Li
VitaMol: Lightweight Molecular Modality Expansion in Discrete Space for Multi-Task Transfer Learning
Yunfan Liu, Zhifeng Gao, Lirong Wu, ..., Chang Yu*, Stan Z. Li
Graphics Capsule: Learning Hierarchical 3D Face Representations from 2D Images
Chang Yu, Xiangyu Zhu, Xiaomei Zhang, Zhaoxiang Zhang, Zhen Lei
HP-Capsule: Unsupervised Face Part Discovery by Hierarchical Parsing Capsule Network
Chang Yu, Xiangyu Zhu, Xiaomei Zhang, Zidu Wang, Zhaoxiang Zhang, Zhen Lei
Out-of-distribution Detection for Reliable Face Recognition
Chang Yu, Xiangyu Zhu, Zhen Lei, Stan Z. Li
Revisiting Marr in Face: A 2D-2.5D-3D Pathway to Part-whole Representation in Neural Networks
Xiangyu Zhu, Chang Yu*, Jiankuo Zhao, Zhaoxiang Zhang, Stan Z. Li, Zhen Lei
Beyond 3DMM: Learning to Capture High-fidelity 3D Face Shape
Xiangyu Zhu, Chang Yu, Di Huang, Zhen Lei, Hao Wang, Stan Z. Li
SteinsGate: Adding Causality to Diffusions for Long Video Generation via Path Integral
Yufei Huang, Liangyu Yuan, ..., Chang Yu*, Stan Z. Li
DaCapo: Score Distillation as Stacked Bridge for Fast and High-quality 3D Editing
Yufei Huang, Bangyan Liao, ..., Chang Yu*, Zhen Lei
π¬ Patents
AI for Science (3 patents pending)
- Siyuan Li, Zicheng Liu, Chang Yu, Lei Xin, Stan Z. Li. An AI-based Method for Integrated Analysis of Multi-Omics Sequence Data.
- Siyuan Li, Lei Xin, Chang Yu, Stan Z. Li. A Method for Integrative Analysis of Multi-Omics Sequence Data.
- Cheng Tan, Zhangyang Gao, Chang Yu, Stan Z. Li. A Protein Structure Representation Method Based on Vector Quantization.
Computer Vision (4 patents granted)
- Xiangyu Zhu, Zhen Lei, Chang Yu. Fine-structure-oriented 3D Face Reconstruction Method, System and Apparatus. ZL202010892658.3
- Zhen Lei, Xiangyu Zhu, Chang Yu. Scene-Adaptive Fine 3D Face Reconstruction Method, System and Electronic Device. ZL202110601213.X
- Zhen Lei, Jianhua Tao, Xiaomei Zhang, Chang Yu. A Capsule-Network-based Facial Expression Recognition Method and Apparatus. ZL202111195200.3
- Xiaomei Zhang, Zhen Lei, Xiangyu Zhu, Chang Yu. Unsupervised Human Body Part Region Determination Method and Apparatus. ZL202311461061.3
π Education
Postdoc, College of Engineering, Westlake University, Hangzhou
Advisor: Prof. Stan Z. Li (IEEE Fellow)
Ph.D. in Computer Application Technology, NLPR, Institute of Automation, Chinese Academy of Sciences (CASIA), Beijing
Advisor: Prof. Zhen Lei (IEEE Fellow)
B.E. in Automation, School of Electronic and Information Engineering, Xi'an Jiaotong University, Xi'an
π Honors and Awards
- 2022 β Pan Deng Scholarship (Second Class), Chinese Academy of Sciences
- 2020 β 2022 β Outstanding Student, University of Chinese Academy of Sciences
- 2018 β Meritorious Winner (M Award), Mathematical Contest in Modeling (MCM/ICM)
- 2017 β National Scholarship
π¬ Academic Service
Conference Reviewer
- CVPR
- NeurIPS
- ICML
- ACM Multimedia
- AAAI
Journal Reviewer
- Pattern Recognition