Split Happens: Combating Advanced Threats with Split Learning and Function Secret Sharing

2025-07-14

AI Summary

Split Learning (SL) faces emerging privacy threats, including model inversion and label inference attacks. Method: This paper proposes a U-shaped Split Learning framework that integrates Function Secret Sharing (FSS) into a U-shaped model partitioning architecture for the first time, enabling fully local label training and significantly enhancing robustness against diverse privacy attacks. The approach synergistically combines FSS, random masking, and a lightweight U-shaped CNN, achieving end-to-end accuracy while drastically reducing communication overhead and training latency. Contribution/Results: Theoretical analysis demonstrates provable security under multiple threat models. Experiments on benchmark datasets show that, compared to a pure-FSS baseline, our method reduces communication cost by ~62% and training time by 57%, with no accuracy loss. This work establishes the first SL paradigm that simultaneously delivers strong privacy guarantees, low computational and communication overhead, and generalizable formal security analysis.

Technology Category

• Machine Learning: Privacy
• Natural Language Processing: Safety and Robustness
• Search and Optimization: Learning to Search

Application Category

• Security and Privacy: Security and privacy of machine learning and AI applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems

Abstract

Split Learning (SL) -- splits a model into two distinct parts to help protect client data while enhancing Machine Learning (ML) processes. Though promising, SL has proven vulnerable to different attacks, thus raising concerns about how effective it may be in terms of data privacy. Recent works have shown promising results for securing SL through the use of a novel paradigm, named Function Secret Sharing (FSS), in which servers obtain shares of a function they compute and operate on a public input hidden with a random mask. However, these works fall short in addressing the rising number of attacks which exist on SL. In SplitHappens, we expand the combination of FSS and SL to U-shaped SL. Similarly to other works, we are able to make use of the benefits of SL by reducing the communication and computational costs of FSS. However, a U-shaped SL provides a higher security guarantee than previous works, allowing a client to keep the labels of the training data secret, without having to share them with the server. Through this, we are able to generalize the security analysis of previous works and expand it to different attack vectors, such as modern model inversion attacks as well as label inference attacks. We tested our approach for two different convolutional neural networks on different datasets. These experiments show the effectiveness of our approach in reducing the training time as well as the communication costs when compared to simply using FSS while matching prior accuracy.

Problem

• Enhancing Split Learning security against advanced threats
• Protecting client data privacy in U-shaped SL models
• Reducing communication and computational costs in FSS

Innovation

• Combining Split Learning with Function Secret Sharing
• U-shaped SL enhances security and privacy
• Reduces communication and computational costs effectively