TT-SEAL: TTD-Aware Selective Encryption for Adversarially-Robust and Low-Latency Edge AI
This work addresses security and latency challenges for edge AI deployments, though it is incremental as it builds on existing selective-encryption and TTD compression methods.
The paper tackled the problem of securing compressed AI models on edge devices by proposing TT-SEAL, a selective-encryption framework for tensor-train decomposed networks, which achieved robustness comparable to full encryption while encrypting only 4.89-15.92% of parameters and reducing AES decryption latency to as low as 2.76% on ResNet-18.
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.