Peer-to-Peer Federated Continual Learning for Naturalistic Driving Action Recognition

TL;DR

This paper introduces FedPC, a peer-to-peer federated learning framework with continual learning for naturalistic driving action recognition, enhancing privacy, efficiency, and personalization without relying on centralized servers.

Contribution

The paper proposes a novel serverless P2P federated learning framework with continual learning for NDAR, improving privacy, efficiency, and adaptability over traditional client-server methods.

Findings

FedPC achieves comparable or better accuracy than traditional FL methods.

FedPC reduces communication, computational, and storage overheads.

FedPC demonstrates strong performance on real-world NDAR datasets.

Abstract

Naturalistic driving action recognition (NDAR) has proven to be an effective method for detecting driver distraction and reducing the risk of traffic accidents. However, the intrusive design of in-cabin cameras raises concerns about driver privacy. To address this issue, we propose a novel peer-to-peer (P2P) federated learning (FL) framework with continual learning, namely FedPC, which ensures privacy and enhances learning efficiency while reducing communication, computational, and storage overheads. Our framework focuses on addressing the clients' objectives within a serverless FL framework, with the goal of delivering personalized and accurate NDAR models. We demonstrate and evaluate the performance of FedPC on two real-world NDAR datasets, including the State Farm Distracted Driver Detection and Track 3 NDAR dataset in the 2023 AICity Challenge. The results of our experiments highlight the strong competitiveness of FedPC compared to the conventional client-to-server (C2S) FLs in terms of performance, knowledge dissemination rate, and compatibility with new clients.