Differentially Private CutMix for Split Learning with Vision Transformer

• Main Authors: Oh, Seungeun, Park, Jihong, Baek, Sihun, Nam, Hyelin, Vepakomma, Praneeth, Raskar, Ramesh, Bennis, Mehdi, Kim, Seong-Lyun
• Format: Journal Article
• Language: English
• Published: 28.10.2022
• Subjects: Computer Science - Computer Vision and Pattern Recognition; Computer Science - Distributed, Parallel, and Cluster Computing; Computer Science - Learning
• DOI: 10.48550/arxiv.2210.15986

Recently, vision transformer (ViT) has started to outpace the conventional CNN in computer vision tasks. Considering privacy-preserving distributed learning with ViT, federated learning (FL) communicates models, which becomes ill-suited due to ViT' s large model size and computing costs. Split learning (SL) detours this by communicating smashed data at a cut-layer, yet suffers from data privacy leakage and large communication costs caused by high similarity between ViT' s smashed data and input data. Motivated by this problem, we propose DP-CutMixSL, a differentially private (DP) SL framework by developing DP patch-level randomized CutMix (DP-CutMix), a novel privacy-preserving inter-client interpolation scheme that replaces randomly selected patches in smashed data. By experiment, we show that DP-CutMixSL not only boosts privacy guarantees and communication efficiency, but also achieves higher accuracy than its Vanilla SL counterpart. Theoretically, we analyze that DP-CutMix amplifies R\'enyi DP (RDP), which is upper-bounded by its Vanilla Mixup counterpart.