---
title: "Exploiting Label Skewness for Spiking Neural Networks in Federated Learning"
authors:
  - "Di Yu"
  - "Xin Du"
  - "Linshan Jiang"
  - "Huijing Zhang"
  - "Shuiguang Deng"
doi: "10.24963/ijcai.2025/767"
url: "https://ijcai.org/proceedings/2025/767"
pdf: "https://ijcai.org/proceedings/2025/0767.pdf"
venue: "Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence"
year: 2025
pages: "6895-6903"
track: "Main Track"
keywords:
  - "Computer Vision: CV: Applications and Systems"
  - "Machine Learning: ML: Federated learning"
  - "Data Mining: DM: Class imbalance and unequal cost"
  - "Humans and AI: HAI: Brain sciences"
---

# Exploiting Label Skewness for Spiking Neural Networks in Federated Learning

Di Yu, Xin Du, Linshan Jiang, Huijing Zhang, Shuiguang Deng

*Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence*, pages 6895-6903. DOI: [10.24963/ijcai.2025/767](https://doi.org/10.24963/ijcai.2025/767)

## Abstract

The energy efficiency of deep spiking neural networks (SNNs) aligns with the constraints of resource-limited edge devices, positioning SNNs as a promising foundation for intelligent applications leveraging the extensive data collected by these devices. To safeguard data privacy, federated learning (FL) facilitates collaborative SNN-based model training by leveraging data distributed across edge devices without transmitting local data to a central server. However, existing FL approaches encounter challenges in handling label-skewed data across devices, inducing drift in the local SNN model and consequently impairing the performance of the global SNN model. To tackle these problems, we propose a novel framework called FedLEC, which incorporates intra-client label weight calibration to balance the learning intensity across local labels and inter-client knowledge distillation to mitigate local SNN model bias caused by label absence. Extensive experiments with three different structured SNNs across five datasets (i.e., three non-neuromorphic and two neuromorphic datasets) demonstrate the efficiency of FedLEC. Compared to seven state-of-the-art FL algorithms, FedLEC achieves an average accuracy improvement of approximately 11.59% for the global SNN model under various label skew distribution settings.