Privacy-Preserving Federated Action Recognition via Differentially Private Selective Tuning and Efficient Communication

TL;DR

This paper introduces FedDP-STECAR, a federated learning framework for video action recognition that enhances privacy through selective tuning and differential privacy, significantly reducing communication costs and maintaining high accuracy.

Contribution

The paper proposes a novel federated learning method that selectively fine-tunes and perturbs only key layers, reducing privacy risks and communication overhead while achieving high accuracy.

Findings

Over 99% reduction in communication traffic.

Up to 70.2% accuracy improvement under strict privacy.

48% faster training with 73.1% accuracy in federated setup.

Abstract

Federated video action recognition enables collaborative model training without sharing raw video data, yet remains vulnerable to two key challenges: \textit{model exposure} and \textit{communication overhead}. Gradients exchanged between clients and the server can leak private motion patterns, while full-model synchronization of high-dimensional video networks causes significant bandwidth and communication costs. To address these issues, we propose \textit{Federated Differential Privacy with Selective Tuning and Efficient Communication for Action Recognition}, namely \textit{FedDP-STECAR}. Our \textit{FedDP-STECAR} framework selectively fine-tunes and perturbs only a small subset of task-relevant layers under Differential Privacy (DP), reducing the surface of information leakage while preserving temporal coherence in video features. By transmitting only the tuned layers during aggregation, communication traffic is reduced by over 99\% compared to full-model updates. Experiments on the UCF-101 dataset using the MViT-B-16x4 transformer show that \textit{FedDP-STECAR} achieves up to \textbf{70.2\% higher accuracy} under strict privacy ($\epsilon=0.65$) in centralized settings and \textbf{48\% faster training} with \textbf{73.1\% accuracy} in federated setups, enabling scalable and privacy-preserving video action recognition. Code available at https://github.com/izakariyya/mvit-federated-videodp