Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FedKLPR: KL-Guided Pruning-Aware Federated Learning for Person Re-Identification

Po-Hsien Yu, Yu-Syuan Tseng, Shao-Yi Chien

Abstract

Person re-identification (re-ID) is a fundamental task in intelligent surveillance and public safety. Federated learning (FL) provides a privacy-preserving paradigm by enabling collaborative model training without centralized data collection. However, applying FL to real-world re-ID systems remains challenging due to two major issues: statistical heterogeneity across clients caused by non-IID data distributions and substantial communication overhead resulting from the frequent transmission of large-scale models. To address these challenges, we propose FedKLPR, a lightweight and communication-efficient federated learning framework for person re-ID. FedKLPR consists of three key components. First, the KL-Divergence Regularization Loss (KLL) constrains local updates by reducing the discrepancy between local and global feature distributions, thereby alleviating the effects of statistical heterogeneity and improving convergence stability under non-IID settings. Second, KL-Divergence-Prune Weighted Aggregation (KLPWA) incorporates both pruning ratio and distributional similarity into the aggregation process, enabling more effective aggregation of pruned local models under non-IID data distributions and enhancing the robustness of the global model. Third, Cross-Round Recovery (CRR) employs a dynamic pruning control mechanism to prevent excessive pruning and preserve model accuracy during iterative compression. Experimental results on eight benchmark datasets demonstrate that FedKLPR achieves substantial communication savings while maintaining competitive accuracy. Compared with state-of-the-art methods, FedKLPR reduces communication cost by 40\%--42\% on ResNet-50 while achieving superior overall performance.

FedKLPR: KL-Guided Pruning-Aware Federated Learning for Person Re-Identification

Abstract

Person re-identification (re-ID) is a fundamental task in intelligent surveillance and public safety. Federated learning (FL) provides a privacy-preserving paradigm by enabling collaborative model training without centralized data collection. However, applying FL to real-world re-ID systems remains challenging due to two major issues: statistical heterogeneity across clients caused by non-IID data distributions and substantial communication overhead resulting from the frequent transmission of large-scale models. To address these challenges, we propose FedKLPR, a lightweight and communication-efficient federated learning framework for person re-ID. FedKLPR consists of three key components. First, the KL-Divergence Regularization Loss (KLL) constrains local updates by reducing the discrepancy between local and global feature distributions, thereby alleviating the effects of statistical heterogeneity and improving convergence stability under non-IID settings. Second, KL-Divergence-Prune Weighted Aggregation (KLPWA) incorporates both pruning ratio and distributional similarity into the aggregation process, enabling more effective aggregation of pruned local models under non-IID data distributions and enhancing the robustness of the global model. Third, Cross-Round Recovery (CRR) employs a dynamic pruning control mechanism to prevent excessive pruning and preserve model accuracy during iterative compression. Experimental results on eight benchmark datasets demonstrate that FedKLPR achieves substantial communication savings while maintaining competitive accuracy. Compared with state-of-the-art methods, FedKLPR reduces communication cost by 40\%--42\% on ResNet-50 while achieving superior overall performance.

Paper Structure

This paper contains 17 sections, 15 equations, 3 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Unsupervised Federated Person re-ID
  4. Pruning in Federated Learning
  5. Methodology
  6. Baseline Model
  7. Overall Architecture
  8. Local Training
  9. Pruning Aggregation-weight Computation
  10. Cross-Round Recovery (CRR)
  11. Experimental Results
  12. Experimental Setup
  13. Performance Evaluation
  14. Pruning Evaluation
  15. Comparison with Various Federated Learning Methods
  16. ...and 2 more sections

Figures (3)

  • Figure 1: Two major challenges in unsupervised federated learning re-ID systems: non-IID data variances across clients and the communication cost between clients and the cloud server.
  • Figure 2: Overview of the FedKLPR framework, consisting of a cloud server and eight clients; each client performs local training with KLL, applies unstructured pruning with CRR, computes KLAW, and uploads the local model, pruning mask, pruning ratio, and KLAW to the cloud, where PRAW is calculated. The cloud combines KLAW and PRAW to aggregate the local models.
  • Figure 3: The CRR mechanism with two-stage verification: Stage 1 checks whether the training accuracy surpasses the target accuracy $Acc_{th}$ and the improvement over previous rounds is within $\Delta_{rd}$; Stage 2 ensures the accuracy drop before and after pruning is less than $\Delta_{ep}$; pruning is executed only if both conditions are satisfied.