针对压缩模型设计轻量加密，兼顾安全与低延迟
TT-SEAL: TTD-Aware Selective Encryption for Adversarially-Robust and Low-Latency Edge AI
- 按敏感度筛选关键张量核，仅加密4.89%-15.92%参数
- 在FPGA上实现，使解密占比从58%降至2.76%
- 适合资源受限的边缘AI部署场景
云端协同的边缘AI需在设备预算紧张下兼顾模型压缩与安全。虽然张量列车分解（TTD）能缩小本地模型，但以往选择性加密研究多假设权重为密集分布，未考虑TTD压缩下的实际可行性。本文提出TT-SEAL框架，用于对TT分解网络进行选择性加密：基于敏感度指标对张量核排序，设定一次性的鲁棒性阈值，并使用值差分隐私优化器，仅用AES加密最少的关键核。在面向迁移攻击的威胁模型下（于FPGA原型边缘处理器上测试），TT-SEAL在保持全加密黑盒同等鲁棒性的同时，仅加密了ResNet-18、MobileNetV2和VGG-16中4.89%-15.92%的参数，将端到端延迟中的AES解密占比降至个位数（如ResNet-18从58%降至2.76%），实现安全且低延迟的边缘AI。
原文摘要 · Abstract (English)
Cloud-edge AI must jointly satisfy model compression and security under tight device budgets. While Tensor-Train Decomposition (TTD) shrinks on-device models, prior selective-encryption studies largely assume dense weights, leaving its practicality under TTD compression unclear. We present TT-SEAL, a selective-encryption framework for TT-decomposed networks. TT-SEAL ranks TT cores with a sensitivity-based importance metric, calibrates a one-time robustness threshold, and uses a value-DP optimizer to encrypt the minimum set of critical cores with AES. Under TTD-aware, transfer-based threat models (and on an FPGA-prototyped edge processor) TT-SEAL matches the robustness of full (black-box) encryption while encrypting as little as 4.89-15.92% of parameters across ResNet-18, MobileNetV2, and VGG-16, and drives the share of AES decryption in end-to-end latency to low single digits (e.g., 58% -> 2.76% on ResNet-18), enabling secure, low-latency edge AI.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。