FedRAV: Hierarchically Federated Region-Learning for Traffic Object Classification of Autonomous Vehicles

TL;DR

This paper proposes a novel hierarchically Federated Region-learning framework of Autonomous Vehicles (FedRAV), a two-stage framework, which adaptively divides a large area containing vehicles into sub-regions based on the defined region-wise distance, and achieves personalized vehicular models and regional models.

Abstract

The emerging federated learning enables distributed autonomous vehicles to train equipped deep learning models collaboratively without exposing their raw data, providing great potential for utilizing explosively growing autonomous driving data. However, considering the complicated traffic environments and driving scenarios, deploying federated learning for autonomous vehicles is inevitably challenged by non-independent and identically distributed (Non-IID) data of vehicles, which may lead to failed convergence and low training accuracy. In this paper, we propose a novel hierarchically Federated Region-learning framework of Autonomous Vehicles (FedRAV), a two-stage framework, which adaptively divides a large area containing vehicles into sub-regions based on the defined region-wise distance, and achieves personalized vehicular models and regional models. This approach ensures that the personalized vehicular model adopts the beneficial models while discarding the unprofitable ones. We validate our FedRAV framework against existing federated learning algorithms on three real-world autonomous driving datasets in various heterogeneous settings. The experiment results demonstrate that our framework outperforms those known algorithms, and improves the accuracy by at least 3.69%. The source code of FedRAV is available at: https://github.com/yjzhai-cs/FedRAV.

Figures (4)

• Figure 1: An example of regional similarities for the samples of Cityscapes on the map. (a) Traffic lights, pedestrians, and cars are dominant in certain-size regions, respectively. (b) All their binary combinations aggregate in certain-size regions. (c) There exist few objects or many objects in some regions. (d) The Significant difference in label distribution amongst the cities.
• Figure 2: Illustration of partitioning mechanism. ① Collect data; ② upload coordinate $V$ and feature vector $L$; ③ divide areas into regions; ④ download regional structure $\{\mathcal{A}_k\}^K_{k=1}$.
• Figure 3: An overview of FedRAV Framework. The main operations consist of: ① local training on private dataset $\mathcal{D}$ ; ② upload model update $\Delta \boldsymbol{w}$; ③ model personalization; ④ intra-region model aggregation; ⑤ download personalized model $\boldsymbol{\tilde{w}}$.
• Figure 4: Comparison of different FL approaches.