De Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
De Li has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Jinyan Wang
Guangxi Normal University
5 shared publications - Xianxian Li
Guangxi Normal University, Guilin, China
5 shared publications - Zeming Gan
School of Computer Science and Engineering, Guangxi Normal University, Guilin, China
4 shared publications - Ximeng Liu
College of Computer and Data Science, Fuzhou University, Fuzhou, China
3 shared publications - Zhou Tan
College of Computer and Data Science, Fuzhou University, Fuzhou, China
3 shared publications - Bin Qu
School of Computer and Artificial Intelligence, Huanghuai University, Zhumadian, Henan, China
2 shared publications
Latest Publications
- 29 May 2026
- Published by Elsevier in Neural Networks
- Vol. 203, 109193
FedEBM: Robust graph federated learning via energy-based model
- 29 May 2026
- Published by Elsevier in Neural Networks
- Vol. 203, 109193
FedEBM: Robust graph federated learning via energy-based model
- 9 May 2026
- Published by Elsevier in Knowledge-Based Systems
- Vol. 346
FedLDP: Model-heterogeneous Federated Graph Learning via Logits Decoupling and Personalization
- 9 May 2026
- Published by Elsevier in Knowledge-Based Systems
- Vol. 346
FedLDP: Model-heterogeneous Federated Graph Learning via Logits Decoupling and Personalization
- 10 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Big Data
- Vol. 12 (2) , 509-524
Toward Federated Learning Against Noisy Clients via CLIP-Guided Prototypes
Federated Noisy Labels Learning (FNLL) allows global model to be jointly trained on multiple clients with varying degrees of noisy labels while preserving privacy, and despite recent research advances, distinguishing between client clean and noisy samples is still tricky since the distribution of labels among clients is always both noisy and class-imbalanced, leading to the poor performance of existing FNLL methods. To address this problem, we propose a novel framework called FedPN, the first framework to utilize Contrastive Language-Image Pre-training (CLIP) for federated noisy labels tasks. Then, to achieve higher performance for the global model, we introduce an attention-based Prototype Adapter to identify more plausible local data for local model training, further improving training stability. We validate the effectiveness of FedPN by conducting extensive experiments on benchmark datasets under both Independently and Identically Distributed (IID) and Non-IID data partitions. The experimental results show that FedPN can effectively filter noisy samples from different clients, and compared with the state-of-the-art FNLL method, the FedPN achieves at most and at least 8.39% and 0.88% performance improvement in the case of highly heterogeneous noisy labels.
- 10 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Big Data
- Vol. 12 (2) , 509-524
Toward Federated Learning Against Noisy Clients via CLIP-Guided Prototypes
Federated Noisy Labels Learning (FNLL) allows global model to be jointly trained on multiple clients with varying degrees of noisy labels while preserving privacy, and despite recent research advances, distinguishing between client clean and noisy samples is still tricky since the distribution of labels among clients is always both noisy and class-imbalanced, leading to the poor performance of existing FNLL methods. To address this problem, we propose a novel framework called FedPN, the first framework to utilize Contrastive Language-Image Pre-training (CLIP) for federated noisy labels tasks. Then, to achieve higher performance for the global model, we introduce an attention-based Prototype Adapter to identify more plausible local data for local model training, further improving training stability. We validate the effectiveness of FedPN by conducting extensive experiments on benchmark datasets under both Independently and Identically Distributed (IID) and Non-IID data partitions. The experimental results show that FedPN can effectively filter noisy samples from different clients, and compared with the state-of-the-art FNLL method, the FedPN achieves at most and at least 8.39% and 0.88% performance improvement in the case of highly heterogeneous noisy labels.
- Open Access
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 7045-7054
FeatShield: Isolating Malicious Feature Extractors for Backdoor-Robust Federated Learning
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 7045-7054
Open AccessFeatShield: Isolating Malicious Feature Extractors for Backdoor-Robust Federated Learning
- Open Access
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 6938-6947
FedRog: Robust Federated Graph Classification for Strong Heterogeneity and High-Noise Scenarios
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 6938-6947
Open AccessFedRog: Robust Federated Graph Classification for Strong Heterogeneity and High-Noise Scenarios
- 10 December 2024
- Published by Elsevier in Neural Networks
- Vol. 184, 107016
FedPD: Defending federated prototype learning against backdoor attacks
- 10 December 2024
- Published by Elsevier in Neural Networks
- Vol. 184, 107016
FedPD: Defending federated prototype learning against backdoor attacks
- 26 November 2024
- Published by Elsevier in Neural Networks
- Vol. 183, 106938
D4A: An efficient and effective defense across agnostic adversarial attacks
- 26 November 2024
- Published by Elsevier in Neural Networks
- Vol. 183, 106938
D4A: An efficient and effective defense across agnostic adversarial attacks
- 20 November 2024
- Published by Elsevier in Neural Networks
- Vol. 182, 106919
Rethinking the impact of noisy labels in graph classification: A utility and privacy perspective
- 20 November 2024
- Published by Elsevier in Neural Networks
- Vol. 182, 106919
Rethinking the impact of noisy labels in graph classification: A utility and privacy perspective
- 13 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (6) , 9955-9968
Optimizing Dynamic Cache Allocation in Vehicular Edge Networks: A Method Combining Multisource Data Prediction and Deep Reinforcement Learning
Currently, most studies on content caching strategies for vehicular networks rely on the Zipf distribution model of content popularity. This approach, however, often fails to accommodate the time-varying nature of content popularity and vehicular traffic. Although some studies have begun considering time-varying content popularity, they overlook the implications of time-varying vehicle traffic, which significantly influences content allocation caching policies and could potentially reduce edge server rental costs for service providers. Addressing this gap, our study takes both time-varying content popularity and vehicle traffic data into account. From the perspective of content service providers, we need to strike a tradeoff in optimizing the hit rate (HR) based on the foundation of utility enhancement. Accordingly, we model the problem as a multiobjective issue that jointly optimizes HR, utility, and replacement cost. By employing a scalar function, we transform this multiobjective problem into a single-objective one, allowing for the modulation of optimization objective priorities via weight tuning. We employ two gated recurrent unit networks to predict both vehicle traffic volume and content popularity for the next moment. We use the predicted area traffic flow volume and content popularity as the system state for our cache allocation decision optimization, which is accomplished using a deep reinforcement learning model. We conducted a series of evaluation experiments using real data sets, and our proposed approach significantly outperforms the baseline scheme.
- 13 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (6) , 9955-9968
Optimizing Dynamic Cache Allocation in Vehicular Edge Networks: A Method Combining Multisource Data Prediction and Deep Reinforcement Learning
Currently, most studies on content caching strategies for vehicular networks rely on the Zipf distribution model of content popularity. This approach, however, often fails to accommodate the time-varying nature of content popularity and vehicular traffic. Although some studies have begun considering time-varying content popularity, they overlook the implications of time-varying vehicle traffic, which significantly influences content allocation caching policies and could potentially reduce edge server rental costs for service providers. Addressing this gap, our study takes both time-varying content popularity and vehicle traffic data into account. From the perspective of content service providers, we need to strike a tradeoff in optimizing the hit rate (HR) based on the foundation of utility enhancement. Accordingly, we model the problem as a multiobjective issue that jointly optimizes HR, utility, and replacement cost. By employing a scalar function, we transform this multiobjective problem into a single-objective one, allowing for the modulation of optimization objective priorities via weight tuning. We employ two gated recurrent unit networks to predict both vehicle traffic volume and content popularity for the next moment. We use the predicted area traffic flow volume and content popularity as the system state for our cache allocation decision optimization, which is accomplished using a deep reinforcement learning model. We conducted a series of evaluation experiments using real data sets, and our proposed approach significantly outperforms the baseline scheme.