Education & Experience
Honors & Awards
I am a Research Fellow at Nanyang Technological University, advised by Prof. Baosheng Yu. I received my Ph.D. from Sichuan University in 2025, advised by Prof. Yi Zhang. From 2024 to 2025, I was a research intern at A*STAR, advised by Joey Tianyi Zhou.
I have published over 20 first-authored papers in top-tier journals and conferences, including CVPR, ICLR, AAAI, IJCAI, IJCV, IEEE T-IFS, IEEE T-NNLS, IEEE T-SMCS, IEEE T-AI, and IEEE J-BHI. citations 1316 One paper was selected as an ESI highly cited paper. If you are interested in academic collaborations, please feel free to contact me.
Research keywords: AI Security, AI in Healthcare, Biometrics, Federated Learning, AI Efficiency.
Please feel free to reach out if you're interested in exploring ideas together, I am always happy to discuss new ideas and explore potential collaborations.
News
Security and Privacy Analysis
LURE: Latent Space Unblocking for Multi-Concept Reawakening in Diffusion Models
Mengyu Sun, Ziyuan Yang†, Andrew Beng Jin Teoh, Junxu Liu†, Haibo Hu, Yi Zhang
International Joint Conference on Artificial Intelligence (IJCAI)
LURE studies multi-concept reawakening in diffusion models and introduces a latent-space unblocking mechanism to recover targeted concepts while retaining controllability.
# AI Security # Generative Models # Trustworthy AI
LURE: Latent Space Unblocking for Multi-Concept Reawakening in Diffusion Models
Mengyu Sun, Ziyuan Yang†, Andrew Beng Jin Teoh, Junxu Liu†, Haibo Hu, Yi Zhang
International Joint Conference on Artificial Intelligence (IJCAI)
LURE studies multi-concept reawakening in diffusion models and introduces a latent-space unblocking mechanism to recover targeted concepts while retaining controllability.
# AI Security # Generative Models # Trustworthy AI
Zero-Sacrifice Persistent-Robustness Adversarial Defense for Pre-Trained Encoders
Zhuxin Lei, Ziyuan Yang†, Yi Zhang†
International Conference on Learning Representations (ICLR)
This paper proposes a persistent-robustness adversarial defense for pre-trained encoders that improves robustness without sacrificing clean-task utility.
# AI Security # Adversarial Defense # Robust Learning
Zero-Sacrifice Persistent-Robustness Adversarial Defense for Pre-Trained Encoders
Zhuxin Lei, Ziyuan Yang†, Yi Zhang†
International Conference on Learning Representations (ICLR)
This paper proposes a persistent-robustness adversarial defense for pre-trained encoders that improves robustness without sacrificing clean-task utility.
# AI Security # Adversarial Defense # Robust Learning
Poisoned Distillation: Injecting Backdoors Into Distilled Datasets Without Raw Data Access
Ziyuan Yang, Ming Yan, Yi Zhang†, Joey Tianyi Zhou
AAAI Conference on Artificial Intelligence (AAAI)
This work studies how backdoors can be injected into distilled datasets without access to the original raw training data, revealing new security risks for dataset distillation pipelines.
# AI Security # Backdoor Attack # Trustworthy AI
Poisoned Distillation: Injecting Backdoors Into Distilled Datasets Without Raw Data Access
Ziyuan Yang, Ming Yan, Yi Zhang†, Joey Tianyi Zhou
AAAI Conference on Artificial Intelligence (AAAI)
This work studies how backdoors can be injected into distilled datasets without access to the original raw training data, revealing new security risks for dataset distillation pipelines.
# AI Security # Backdoor Attack # Trustworthy AI
A Dual-Level Cancelable Framework for Palmprint Verification and Hack-Proof Data Storage
Ziyuan Yang, Ming Kang, Andrew Beng Jin Teoh
IEEE Transactions on Information Forensics and Security (IEEE T-IFS)
This work proposes a dual-level cancelable biometric framework for secure palmprint verification and hack-proof template storage.
# Biometrics # Privacy-Preserving Learning # AI Security
A Dual-Level Cancelable Framework for Palmprint Verification and Hack-Proof Data Storage
Ziyuan Yang, Ming Kang, Andrew Beng Jin Teoh
IEEE Transactions on Information Forensics and Security (IEEE T-IFS)
This work proposes a dual-level cancelable biometric framework for secure palmprint verification and hack-proof template storage.
# Biometrics # Privacy-Preserving Learning # AI Security
A Novel Privacy-Enhancing Framework for Low-Dose CT Denoising
Ziyuan Yang, Huijie Huangfu, Maosong Ran, Zhiwen Wang, Hui Yu, Mengyu Sun, Yi Zhang
IEEE Transactions on Artificial Intelligence (IEEE T-AI)
This paper presents a privacy-enhancing framework for low-dose CT denoising that preserves sensitive information while maintaining reconstruction quality.
# AI in Healthcare # Privacy-Preserving Learning # Medical Imaging
A Novel Privacy-Enhancing Framework for Low-Dose CT Denoising
Ziyuan Yang, Huijie Huangfu, Maosong Ran, Zhiwen Wang, Hui Yu, Mengyu Sun, Yi Zhang
IEEE Transactions on Artificial Intelligence (IEEE T-AI)
This paper presents a privacy-enhancing framework for low-dose CT denoising that preserves sensitive information while maintaining reconstruction quality.
# AI in Healthcare # Privacy-Preserving Learning # Medical Imaging
Inject Backdoor in Measured Data to Jeopardize Full-Stack Medical Image Analysis System
Ziyuan Yang, Yingyu Chen, Mengyu Sun, Yi Zhang†
International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI)
This paper presents a novel pre-imaging backdoor attack called LTGM that injects learnable triggers into measured medical data to compromise downstream image analysis tasks without affecting reconstruction quality, exposing vulnerabilities in full-stack medical image analysis systems.
# Medical Image Analysis # Trustworthy AI
Inject Backdoor in Measured Data to Jeopardize Full-Stack Medical Image Analysis System
Ziyuan Yang, Yingyu Chen, Mengyu Sun, Yi Zhang†
International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI)
This paper presents a novel pre-imaging backdoor attack called LTGM that injects learnable triggers into measured medical data to compromise downstream image analysis tasks without affecting reconstruction quality, exposing vulnerabilities in full-stack medical image analysis systems.
# Medical Image Analysis # Trustworthy AI
All Research
Federated Learning
Patient-Level Anatomy Meets Scanning-Level Physics: Personalized Federated Low-Dose CT Denoising Empowered by Large Language Model
Ziyuan Yang, Yingyu Chen, Zhiwen Wang, Hongming Shan, Yang Chen, Yi Zhang†
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
This paper introduces SCAN-PhysFed, a personalized federated learning framework for low-dose CT denoising that leverages scanning- and anatomy-level physics-informed prompts, guided by a medical large language model, to achieve robust and generalizable reconstruction across diverse scanning protocols while preserving patient privacy.
# Federated Learning # Medical Imaging # LLM
Patient-Level Anatomy Meets Scanning-Level Physics: Personalized Federated Low-Dose CT Denoising Empowered by Large Language Model
Ziyuan Yang, Yingyu Chen, Zhiwen Wang, Hongming Shan, Yang Chen, Yi Zhang†
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
This paper introduces SCAN-PhysFed, a personalized federated learning framework for low-dose CT denoising that leverages scanning- and anatomy-level physics-informed prompts, guided by a medical large language model, to achieve robust and generalizable reconstruction across diverse scanning protocols while preserving patient privacy.
# Federated Learning # Medical Imaging # LLM
Enhancing Federated Learning through Exploring Filter-Aware Relationships and Personalizing Local Structures
Ziyuan Yang, Zerui Shao, Hui Yu, Huijie Huangfu, Andrew Beng Jin Teoh, Xiaoxiao Li, Hongming Shan, Yi Zhang
Pattern Recognition (PR)
This paper improves federated learning by modeling filter-aware relationships and personalizing local model structures for stronger cross-domain generalization.
# Federated Learning # Biometrics
Enhancing Federated Learning through Exploring Filter-Aware Relationships and Personalizing Local Structures
Ziyuan Yang, Zerui Shao, Hui Yu, Huijie Huangfu, Andrew Beng Jin Teoh, Xiaoxiao Li, Hongming Shan, Yi Zhang
Pattern Recognition (PR)
This paper improves federated learning by modeling filter-aware relationships and personalizing local model structures for stronger cross-domain generalization.
# Federated Learning # Biometrics
Hypernetwork-Based Physics-Driven Personalized Federated Learning for CT Imaging
Ziyuan Yang, Wenjun Xia, Wenjun Xia, Yingyu Chen, Xiaoxiao Li, Yi Zhang†
IEEE Transactions on Neural Networks and Learning Systems (IEEE T-NNLS)
HyperFed is a physics-driven personalized federated learning framework for CT imaging that uses hypernetworks to model institution-specific characteristics while preserving privacy and reconstruction quality across sites.
# Federated Learning # Medical Imaging
Hypernetwork-Based Physics-Driven Personalized Federated Learning for CT Imaging
Ziyuan Yang, Wenjun Xia, Wenjun Xia, Yingyu Chen, Xiaoxiao Li, Yi Zhang†
IEEE Transactions on Neural Networks and Learning Systems (IEEE T-NNLS)
HyperFed is a physics-driven personalized federated learning framework for CT imaging that uses hypernetworks to model institution-specific characteristics while preserving privacy and reconstruction quality across sites.
# Federated Learning # Medical Imaging
Dynamic Corrected Split Federated Learning with Homomorphic Encryption for U-Shaped Medical Image Networks
Ziyuan Yang, Yingyu Chen, Huijie Huangfu, Maosong Ran, Hui Wang, Xiaoxiao Li, Yi Zhang†
IEEE Journal of Biomedical and Health Informatics (IEEE JBHI)
This paper proposes DC-SFL, a hybrid split-federated learning framework for U-shaped medical image networks that combines dynamic weight correction and homomorphic encryption to ensure data and model privacy, stabilize training under heterogeneous data, and achieve competitive performance across medical imaging tasks.
# Federated Learning # Medical Learning
Dynamic Corrected Split Federated Learning with Homomorphic Encryption for U-Shaped Medical Image Networks
Ziyuan Yang, Yingyu Chen, Huijie Huangfu, Maosong Ran, Hui Wang, Xiaoxiao Li, Yi Zhang†
IEEE Journal of Biomedical and Health Informatics (IEEE JBHI)
This paper proposes DC-SFL, a hybrid split-federated learning framework for U-shaped medical image networks that combines dynamic weight correction and homomorphic encryption to ensure data and model privacy, stabilize training under heterogeneous data, and achieve competitive performance across medical imaging tasks.
# Federated Learning # Medical Learning
All Research
Biometrics
FedPalm: a General Federated Learning Framework for Closed-And Open-Set Palmprint Verification
Ziyuan Yang, Yingyu Chen, Chengrui Gao, Andrew Beng Jin Teoh, Bob Zhang, Yi Zhang†
IEEE Transactions on Information Forensics and Security (IEEE T-IFS)
This paper proposes FedPalm, a unified federated learning framework for palmprint verification that combines personalized local textural experts with a shared global expert to achieve robust performance in both closed-set and open-set scenarios while preserving biometric privacy.
# Federated Learning # Biometrics
FedPalm: a General Federated Learning Framework for Closed-And Open-Set Palmprint Verification
Ziyuan Yang, Yingyu Chen, Chengrui Gao, Andrew Beng Jin Teoh, Bob Zhang, Yi Zhang†
IEEE Transactions on Information Forensics and Security (IEEE T-IFS)
This paper proposes FedPalm, a unified federated learning framework for palmprint verification that combines personalized local textural experts with a shared global expert to achieve robust performance in both closed-set and open-set scenarios while preserving biometric privacy.
# Federated Learning # Biometrics
Physics-Driven Spectrum-Consistent Federated Learning for Palmprint Verification
Ziyuan Yang, Andrew Beng Jin Teoh, Bob Zhang, Lu Leng, Yi Zhang
International Journal of Computer Vision (IJCV)
This paper introduces a physics-driven, spectrum-consistent federated learning framework for robust palmprint verification across sensing conditions.
# Federated Learning # Biometrics
Physics-Driven Spectrum-Consistent Federated Learning for Palmprint Verification
Ziyuan Yang, Andrew Beng Jin Teoh, Bob Zhang, Lu Leng, Yi Zhang
International Journal of Computer Vision (IJCV)
This paper introduces a physics-driven, spectrum-consistent federated learning framework for robust palmprint verification across sensing conditions.
# Federated Learning # Biometrics
Comprehensive Competition Mechanism in Palmprint Recognition
Ziyuan Yang, Huijie Huangfu, Lu Leng, Bob Zhang, Andrew Beng Jin Teoh, Yi Zhang
IEEE Transactions on Information Forensics and Security (IEEE T-IFS)
This paper presents a comprehensive competition mechanism for palmprint recognition to improve discriminative representation learning and verification performance.
# Biometrics # Representation Learning
Comprehensive Competition Mechanism in Palmprint Recognition
Ziyuan Yang, Huijie Huangfu, Lu Leng, Bob Zhang, Andrew Beng Jin Teoh, Yi Zhang
IEEE Transactions on Information Forensics and Security (IEEE T-IFS)
This paper presents a comprehensive competition mechanism for palmprint recognition to improve discriminative representation learning and verification performance.
# Biometrics # Representation Learning
Deep Learning in Palmprint Recognition: a Comprehensive Survey
Chengrui Gao, Ziyuan Yang†, Wei Jia, Lu Leng, Bob Zhang, Andrew Beng Jin Teoh
IEEE Transactions on Systems, Man, and Cybernetics: Systems (IEEE T-SMCS)
This survey reviews recent deep learning methods for palmprint recognition, covering datasets, modeling strategies, challenges, and future research directions.
# Biometrics # Survey
Deep Learning in Palmprint Recognition: a Comprehensive Survey
Chengrui Gao, Ziyuan Yang†, Wei Jia, Lu Leng, Bob Zhang, Andrew Beng Jin Teoh
IEEE Transactions on Systems, Man, and Cybernetics: Systems (IEEE T-SMCS)
This survey reviews recent deep learning methods for palmprint recognition, covering datasets, modeling strategies, challenges, and future research directions.
# Biometrics # Survey
All Research
AI in Healthcare
FACT: Fuzzy Alignment with Comorbidity Topology for Reliable Multi-Label Medical Image Diagnosis
Yingyu Chen, Yongqiang Huang, Yang Qin, Ziyuan Yang, Lang Yuan, Maosong Ran, Yi Zhang†
International Conference on Machine Learning (ICML)
This paper proposes FACT, a framework that reinterprets multi-label medical image diagnosis as a fuzzy alignment problem, leveraging vector quantization to construct atomic visual evidence and a graph convolutional network to embed comorbidity topology, with a metric-based fuzzy membership function derived from RKHS theory.
# Medical Analysis # Multi-Label Learning # Vision-Language
FACT: Fuzzy Alignment with Comorbidity Topology for Reliable Multi-Label Medical Image Diagnosis
Yingyu Chen, Yongqiang Huang, Yang Qin, Ziyuan Yang, Lang Yuan, Maosong Ran, Yi Zhang†
International Conference on Machine Learning (ICML)
This paper proposes FACT, a framework that reinterprets multi-label medical image diagnosis as a fuzzy alignment problem, leveraging vector quantization to construct atomic visual evidence and a graph convolutional network to embed comorbidity topology, with a metric-based fuzzy membership function derived from RKHS theory.
# Medical Analysis # Multi-Label Learning # Vision-Language
Double Banking on Knowledge: a Unified All-In-One Framework for Unpaired Multi-Modality Semi-Supervised Medical Image Segmentation
Yingyu Chen, Ziyuan Yang, Zhongzhou Zhang, Ming Yan, Hui Yu, Yan Liu, Yi Zhang†
IEEE Transactions on Biomedical Engineering (IEEE T-BME)
This paper proposes a unified all-in-one framework for unpaired multi-modality semi-supervised medical image segmentation that leverages learnable knowledge banks, modality-adaptive weighting, and dual consistency to capture both modality-invariant and modality-specific features, enabling scalable and robust segmentation across multiple modalities.
# Medical Analysis # Weakly-Supervised Learning
Double Banking on Knowledge: a Unified All-In-One Framework for Unpaired Multi-Modality Semi-Supervised Medical Image Segmentation
Yingyu Chen, Ziyuan Yang, Zhongzhou Zhang, Ming Yan, Hui Yu, Yan Liu, Yi Zhang†
IEEE Transactions on Biomedical Engineering (IEEE T-BME)
This paper proposes a unified all-in-one framework for unpaired multi-modality semi-supervised medical image segmentation that leverages learnable knowledge banks, modality-adaptive weighting, and dual consistency to capture both modality-invariant and modality-specific features, enabling scalable and robust segmentation across multiple modalities.
# Medical Analysis # Weakly-Supervised Learning
Generalizable MRI Motion Correction via Compressed Sensing Equivariant Imaging Prior
Zhiwen Wang, Maosong Ran, Ziyuan Yang
IEEE Transactions on Circuits and Systems for Video Technology (IEEE T-CSVT)
This work proposes an equivariant imaging prior for generalizable MRI motion correction under compressed sensing settings.
# AI in Healthcare # Medical Imaging
Generalizable MRI Motion Correction via Compressed Sensing Equivariant Imaging Prior
Zhiwen Wang, Maosong Ran, Ziyuan Yang
IEEE Transactions on Circuits and Systems for Video Technology (IEEE T-CSVT)
This work proposes an equivariant imaging prior for generalizable MRI motion correction under compressed sensing settings.
# AI in Healthcare # Medical Imaging
SOUL-Net: a Sparse and Low-Rank Unrolling Network for Spectral CT Image Reconstruction
Xiang Chen, Wenjun Xia, Ziyuan Yang
IEEE Transactions on Neural Networks and Learning Systems (IEEE T-NNLS)
SOUL-Net is a sparse and low-rank unrolling network for spectral CT image reconstruction that improves image quality while preserving reconstruction efficiency.
# AI in Healthcare # Medical Imaging
SOUL-Net: a Sparse and Low-Rank Unrolling Network for Spectral CT Image Reconstruction
Xiang Chen, Wenjun Xia, Ziyuan Yang
IEEE Transactions on Neural Networks and Learning Systems (IEEE T-NNLS)
SOUL-Net is a sparse and low-rank unrolling network for spectral CT image reconstruction that improves image quality while preserving reconstruction efficiency.
# AI in Healthcare # Medical Imaging
RegFormer: a Local-Nonlocal Regularization-Based Model for Sparse-View CT Reconstruction
Wenjun Xia, Ziyuan Yang, Zexin Lu
IEEE Transactions on Radiation and Plasma Medical Sciences (IEEE T-RPMS)
RegFormer introduces a local-nonlocal regularization framework for sparse-view CT reconstruction to improve robustness and reconstruction fidelity.
# AI in Healthcare # Medical Imaging
RegFormer: a Local-Nonlocal Regularization-Based Model for Sparse-View CT Reconstruction
Wenjun Xia, Ziyuan Yang, Zexin Lu
IEEE Transactions on Radiation and Plasma Medical Sciences (IEEE T-RPMS)
RegFormer introduces a local-nonlocal regularization framework for sparse-view CT reconstruction to improve robustness and reconstruction fidelity.
# AI in Healthcare # Medical Imaging
All Research
Service
Editor
Guest Editor for CMC - Computers, Materials & Continua.
Guest Editor for Sensors.
Journal Reviewer
IEEE T-PAMI, IJCV, IEEE T-IP, IEEE T-IFS, IEEE T-DSC, IEEE T-MI, IEEE T-MM, IEEE T-NNLS, IEEE T-MC, IEEE T-ASL, IEEE T-CSVT, IEEE T-SMCS, IEEE T-II, IEEE T-BC, IEEE COMST, IEEE TCDS, IEEE T-Mech, IEEE SPL, IEEE IoTJ, IEEE Sens-J, IEEE J-BHI, AIRE, AIME, and IET CV/SP/Biom.
Conference Reviewer
CVPR, NeurIPS, ICCV, AAAI, ACM MM, ECCV, ICME, MICCAI, EMNLP, IJCNN, ISBI, and WCCI.