Deploying Convolutional Networks on Untrusted Platforms Using 2D Holographic Reduced Representations

Mohammad Mahmudul Alam, Edward Raff, Tim Oates, James Holt
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:367-393, 2022.

Abstract

Due to the computational cost of running inference for a neural network, the need to deploy the inferential steps on a third party’s compute environment or hardware is common. If the third party is not fully trusted, it is desirable to obfuscate the nature of the inputs and outputs, so that the third party can not easily determine what specific task is being performed. Provably secure protocols for leveraging an untrusted party exist but are too computational demanding to run in practice. We instead explore a different strategy of fast, heuristic security that we call Connectionist Symbolic Pseudo Secrets. By leveraging Holographic Reduced Representations (HRRs), we create a neural network with a pseudo-encryption style defense that empirically shows robustness to attack, even under threat models that unrealistically favor the adversary.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-alam22a, title = {Deploying Convolutional Networks on Untrusted Platforms Using 2{D} Holographic Reduced Representations}, author = {Alam, Mohammad Mahmudul and Raff, Edward and Oates, Tim and Holt, James}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {367--393}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/alam22a/alam22a.pdf}, url = {https://proceedings.mlr.press/v162/alam22a.html}, abstract = {Due to the computational cost of running inference for a neural network, the need to deploy the inferential steps on a third party’s compute environment or hardware is common. If the third party is not fully trusted, it is desirable to obfuscate the nature of the inputs and outputs, so that the third party can not easily determine what specific task is being performed. Provably secure protocols for leveraging an untrusted party exist but are too computational demanding to run in practice. We instead explore a different strategy of fast, heuristic security that we call Connectionist Symbolic Pseudo Secrets. By leveraging Holographic Reduced Representations (HRRs), we create a neural network with a pseudo-encryption style defense that empirically shows robustness to attack, even under threat models that unrealistically favor the adversary.} }
Endnote
%0 Conference Paper %T Deploying Convolutional Networks on Untrusted Platforms Using 2D Holographic Reduced Representations %A Mohammad Mahmudul Alam %A Edward Raff %A Tim Oates %A James Holt %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-alam22a %I PMLR %P 367--393 %U https://proceedings.mlr.press/v162/alam22a.html %V 162 %X Due to the computational cost of running inference for a neural network, the need to deploy the inferential steps on a third party’s compute environment or hardware is common. If the third party is not fully trusted, it is desirable to obfuscate the nature of the inputs and outputs, so that the third party can not easily determine what specific task is being performed. Provably secure protocols for leveraging an untrusted party exist but are too computational demanding to run in practice. We instead explore a different strategy of fast, heuristic security that we call Connectionist Symbolic Pseudo Secrets. By leveraging Holographic Reduced Representations (HRRs), we create a neural network with a pseudo-encryption style defense that empirically shows robustness to attack, even under threat models that unrealistically favor the adversary.
APA
Alam, M.M., Raff, E., Oates, T. & Holt, J.. (2022). Deploying Convolutional Networks on Untrusted Platforms Using 2D Holographic Reduced Representations. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:367-393 Available from https://proceedings.mlr.press/v162/alam22a.html.

Related Material