Lu Huang

Lu Huang

Auto-generated profile, based on publicly available information

Lu Huang


Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
18 October 2022 in Expert Systems with Applications
Verified
Journal Article
Closed access
JCJunbao Chen
Junbao Chen
JXJingfeng Xue
Jingfeng Xue
YWYong Wang
Yong Wang
LHLu Huang
Lu Huang
Thar BakerThar Baker
Thar Baker
ZZZhixiong Zhou
Zhixiong Zhou

10 June 2024 in Information Sciences
Verified
Journal Article
Closed access
LHLu Huang
Lu Huang
JXJingfeng Xue
Jingfeng Xue
YWYong Wang
Yong Wang
JCJunbao Chen
Junbao Chen
TLTianwei Lei
Tianwei Lei

28 July 2023 in Journal of Information Security and Applications
Verified
Journal Article
Closed access
LHLu Huang
Lu Huang
JXJingfeng Xue
Jingfeng Xue
YWYong Wang
Yong Wang
ZLZhenyan Liu
Zhenyan Liu
JCJunbao Chen
Junbao Chen
ZKZixiao Kong
Zixiao Kong

26 June 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
JXJingfeng Xue
Jingfeng Xue
ZYZuobin Ying
Zuobin Ying
YMYuxi Ma
Yuxi Ma
YWYong Wang
Yong Wang
ZLZhenyan Liu
Zhenyan Liu
JCJunbao Chen
Junbao Chen
Federated domain generalization aims to learn a global model that generalizes well to unseen domains. Feature alignment with prototypes enhances domain-invariant representation learning, aligning inconsistent feature spaces caused by domain skew. However, we identify two key problems that limits feature alignment: local prototypes become unreliable due to mismatched batch normalization statistics and insufficient inter-class separation, while conventional prototype aggregation biases the global feature space toward overrepresented domains. Building on these findings, we propose FedCoda, a cross-domain feature alignment framework that calibrates local prototype geometry and debiases global prototype aggregation. FedCoda first performs calibrated feature alignment by transitioning to global batch normalization statistics in the later training stage, while introducing a prototype calibration loss that encourages well-separated class prototypes. It then applies fair prototype aggregation, which optimizes global prototypes by minimizing the variance of their distances to local prototypes, reducing dominance from domains with more clients. Extensive experiments show that FedCoda consistently improves generalization ability and representation learning under various settings, demonstrating the importance of jointly addressing prototype reliability and aggregation fairness for federated domain generalization.