Jue Xiao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Jue Xiao has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Hewang Nie
Key Lab of Education Blockchain and Intelligent Technology, Ministry of Education, Guangxi Normal University, Guilin, China
5 shared publications - Songfeng Lu
Huazhong University of Science and Technology
4 shared publications - Zhi Lu
Huazhong University of Science and Technology
3 shared publications - Renfei Shen
Huazhong University of Science and Technology
2 shared publications - Xueming Tang
Hubei Key Laboratory of Distributed System Security, Hubei Engineering Research Center on Big Data Security, School of Cyber Science and Engineering, Huazhong University of Science and Technology, Wuhan, China
2 shared publications - Zepu Yi
School of Cyber Science and Engineering, Huazhong University of Science and Technology, Wuhan 430074, China
2 shared publications
Latest Publications
- Open Access
- 11 June 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Privacy and Security
- Vol. 29 (3) , 1-28
Watermarking for Model Ownership Verification:Invisible at Deployment, Activated by Updates
Deep neural networks for image classification require protection against unauthorized use and redistribution. Existing watermarking methods suffer from a critical vulnerability: watermarks are always active and detectable, allowing adversaries to identify and remove them before deployment. We propose DormMark, a novel framework for image classification models that introduces delayed-activation watermarks which remain dormant and hidden under deployment-time black-box query auditing during initial deployment, but automatically activate upon fine-tuning. Our approach employs a three-stage training paradigm: (1) embedding watermarks using triggered samples, (2) masking to suppress watermark functionality while preserving its latent presence, and (3) activation through standard fine-tuning without owner intervention. This mechanism exploits neural networks’ forgetting-remembering behaviors during continued training, creating a fragile equilibrium that behaves similarly to a clean model under deployment-time black-box auditing but reliably manifests ownership indicators after modification. We consider a private-key black-box verification setting in which the owner keeps the concrete trigger instances secret. Experiments across multiple architectures (VGG19, ResNet-18/56, DenseNet-121, WideResNet-34) and datasets (CIFAR-10, CIFAR-100, GTSRB) demonstrate 100% watermark success rates, high imperceptibility (PSNR > 38 dB, SSIM = 0.99), negligible accuracy loss (< 0.04%), and robustness against 80% parameter pruning. DormMark represents a paradigm shift from static to conditionally-activated ownership verification, providing a more robust framework for intellectual property protection.
- 11 June 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Privacy and Security
- Vol. 29 (3) , 1-28
Open AccessWatermarking for Model Ownership Verification:Invisible at Deployment, Activated by Updates
Deep neural networks for image classification require protection against unauthorized use and redistribution. Existing watermarking methods suffer from a critical vulnerability: watermarks are always active and detectable, allowing adversaries to identify and remove them before deployment. We propose DormMark, a novel framework for image classification models that introduces delayed-activation watermarks which remain dormant and hidden under deployment-time black-box query auditing during initial deployment, but automatically activate upon fine-tuning. Our approach employs a three-stage training paradigm: (1) embedding watermarks using triggered samples, (2) masking to suppress watermark functionality while preserving its latent presence, and (3) activation through standard fine-tuning without owner intervention. This mechanism exploits neural networks’ forgetting-remembering behaviors during continued training, creating a fragile equilibrium that behaves similarly to a clean model under deployment-time black-box auditing but reliably manifests ownership indicators after modification. We consider a private-key black-box verification setting in which the owner keeps the concrete trigger instances secret. Experiments across multiple architectures (VGG19, ResNet-18/56, DenseNet-121, WideResNet-34) and datasets (CIFAR-10, CIFAR-100, GTSRB) demonstrate 100% watermark success rates, high imperceptibility (PSNR > 38 dB, SSIM = 0.99), negligible accuracy loss (< 0.04%), and robustness against 80% parameter pruning. DormMark represents a paradigm shift from static to conditionally-activated ownership verification, providing a more robust framework for intellectual property protection.
- Open Access
- 20 March 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Intelligent Systems and Technology
- Vol. 17 (3) , 1-28
Delete but Not Gone: Reactivation of Neural Network Watermarks
As Deep Neural Networks (DNNs) become integral to critical applications, protecting their Intellectual Property (IP) has become paramount. Neural network watermarking is a technique that embeds unique identifiers into models, asserting ownership and deterring unauthorized use. However, sophisticated attacks can deactivate or remove these watermarks without significantly compromising model performance, undermining current protection strategies. In this article, we introduce the first method for reactivating deactivated neural network watermarks in altered DNN models without requiring access to the original model parameters or training data. By formulating the reactivation process as an optimization problem, we employ projected gradient descent to identify new trigger inputs that restore the embedded watermark. Regularization techniques are incorporated to ensure these triggers resemble legitimate inputs, enhancing both stealth and practicality. Through experiments on various benchmark datasets and model architectures, we demonstrate the effectiveness of our method against common model alterations, including fine-tuning, pruning, and surrogate model attacks. Our work addresses a critical gap in DNN IP protection, offering a robust and practical solution for watermark reactivation. This empowers model owners to assert their rights even in the face of advanced adversarial tactics.
- 20 March 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Intelligent Systems and Technology
- Vol. 17 (3) , 1-28
Open AccessDelete but Not Gone: Reactivation of Neural Network Watermarks
As Deep Neural Networks (DNNs) become integral to critical applications, protecting their Intellectual Property (IP) has become paramount. Neural network watermarking is a technique that embeds unique identifiers into models, asserting ownership and deterring unauthorized use. However, sophisticated attacks can deactivate or remove these watermarks without significantly compromising model performance, undermining current protection strategies. In this article, we introduce the first method for reactivating deactivated neural network watermarks in altered DNN models without requiring access to the original model parameters or training data. By formulating the reactivation process as an optimization problem, we employ projected gradient descent to identify new trigger inputs that restore the embedded watermark. Regularization techniques are incorporated to ensure these triggers resemble legitimate inputs, enhancing both stealth and practicality. Through experiments on various benchmark datasets and model architectures, we demonstrate the effectiveness of our method against common model alterations, including fine-tuning, pruning, and surrogate model attacks. Our work addresses a critical gap in DNN IP protection, offering a robust and practical solution for watermark reactivation. This empowers model owners to assert their rights even in the face of advanced adversarial tactics.
- 6 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36 (5) , 6428-6442
Exclusive and Updateable Watermarking for Neural Network Model Trading
Neural-network model trading raises two unmet requirements for watermarking: exclusive ownership verification and updateability (transfer/revoke) without retraining. We present a training-time framework that jointly embeds a keydriven dynamic label-mapping and a proactive, self-defending trigger. The label-mapping encodes an owner-specific signature that authorized key holders can verify, update, or transfer by rotating keys; the defensive trigger actively hardens the model against unauthorized backdoor insertion and fakewatermark attempts. Our design yields three properties: (i) exclusivity—only the correct key decodes the watermark; (ii) updateability—ownership can be changed without retraining the backbone; and (iii) robustness—the watermark persists under common post-deployment changes. Across CIFAR-10/100, GTSRB, and Tiny ImageNet on five architectures, the method achieves (> 97%) watermark success rate while preserving original accuracy within 0.5 percentage points, retains (>85%) watermark strength after fine-tuning and (60%) pruning, and reduces fake-watermark success to (< 1.5%) (vs. (> 95%) on unprotected models). By coupling updateable encoding with proactive defense, our approach offers a practical, scalable path to secure, transferable ownership verification for neural-network marketplaces and model exchanges.
- 6 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36 (5) , 6428-6442
Exclusive and Updateable Watermarking for Neural Network Model Trading
Neural-network model trading raises two unmet requirements for watermarking: exclusive ownership verification and updateability (transfer/revoke) without retraining. We present a training-time framework that jointly embeds a keydriven dynamic label-mapping and a proactive, self-defending trigger. The label-mapping encodes an owner-specific signature that authorized key holders can verify, update, or transfer by rotating keys; the defensive trigger actively hardens the model against unauthorized backdoor insertion and fakewatermark attempts. Our design yields three properties: (i) exclusivity—only the correct key decodes the watermark; (ii) updateability—ownership can be changed without retraining the backbone; and (iii) robustness—the watermark persists under common post-deployment changes. Across CIFAR-10/100, GTSRB, and Tiny ImageNet on five architectures, the method achieves (> 97%) watermark success rate while preserving original accuracy within 0.5 percentage points, retains (>85%) watermark strength after fine-tuning and (60%) pruning, and reduces fake-watermark success to (< 1.5%) (vs. (> 95%) on unprotected models). By coupling updateable encoding with proactive defense, our approach offers a practical, scalable path to secure, transferable ownership verification for neural-network marketplaces and model exchanges.
- 27 January 2025
- Published by Elsevier in Neural Networks
- Vol. 185, 107199
Federated learning with bilateral defense via blockchain
- 27 January 2025
- Published by Elsevier in Neural Networks
- Vol. 185, 107199
Federated learning with bilateral defense via blockchain
- 26 August 2024
- Published by Springer Nature
- p. 332-345
VeriChroma: Ownership Verification for Federated Models via RGB Filters
- 26 August 2024
- Published by Springer Nature
- p. 332-345
VeriChroma: Ownership Verification for Federated Models via RGB Filters
- Open Access
- 18 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Machine Learning in Communications and Networking
- Vol. 2, 150-168
Privacy-Preserving Federated Class-Incremental Learning
Federated Learning (FL) offers a collaborative training framework, aggregating model parameters from decentralized clients. Many existing models, however, assume static, predetermined data classes within FLa frequently unrealistic assumption. Real-time data additions from clients can degrade global model recognition of established classes due to catastrophic forgetting. This is exacerbated when new clients, unfamiliar to previous participants, join sporadically. Additionally, there’s an imperative for client data privacy. Addressing these, we propose the Privacy-Preserving Federated Class-Incremental Learning (PP-FCIL) approach. This methodology ensures content-level privacy and significantly alleviates the risk of catastrophic forgetting in FCIL. To our knowledge, this is the first research seeks to embed differential privacy into the FCIL settings. Specifically, we introduce a dual-model structure that uses adaptive fusion of new and old knowledge to obtain a new global model. We also propose a multi-factor dynamic weighted aggregation strategy that considers several factors, such as data imbalance timeliness of the model, to speed up global model aggregation and accuracy. For privacy protection, we use Bayesian differential privacy to provide more balanced privacy protection for different datasets. Finally, we conducted experiments on CIFAR-100 and ImageNet to compare our method with other methods and verify its superiority.
- 18 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Machine Learning in Communications and Networking
- Vol. 2, 150-168
Open AccessPrivacy-Preserving Federated Class-Incremental Learning
Federated Learning (FL) offers a collaborative training framework, aggregating model parameters from decentralized clients. Many existing models, however, assume static, predetermined data classes within FLa frequently unrealistic assumption. Real-time data additions from clients can degrade global model recognition of established classes due to catastrophic forgetting. This is exacerbated when new clients, unfamiliar to previous participants, join sporadically. Additionally, there’s an imperative for client data privacy. Addressing these, we propose the Privacy-Preserving Federated Class-Incremental Learning (PP-FCIL) approach. This methodology ensures content-level privacy and significantly alleviates the risk of catastrophic forgetting in FCIL. To our knowledge, this is the first research seeks to embed differential privacy into the FCIL settings. Specifically, we introduce a dual-model structure that uses adaptive fusion of new and old knowledge to obtain a new global model. We also propose a multi-factor dynamic weighted aggregation strategy that considers several factors, such as data imbalance timeliness of the model, to speed up global model aggregation and accuracy. For privacy protection, we use Bayesian differential privacy to provide more balanced privacy protection for different datasets. Finally, we conducted experiments on CIFAR-100 and ImageNet to compare our method with other methods and verify its superiority.