Preprint · 2019 · Open access

Privacy-Preserving Inference in Machine Learning Services Using Trusted Execution Environments

Krishna Giri Narra, Zhifeng Lin, Yongqin Wang, Keshav Balasubramaniam, Murali Annavaram

arXiv (Cornell University) · Published 7 December 2019 · DOI 10.48550/arxiv.1912.03485

Open PDFView at the publisherCite this paper

In brief

“Privacy-Preserving Inference in Machine Learning Services Using Trusted Execution Environments” is a 2019 preprint by Krishna Giri Narra and 4 co-authors, published in arXiv (Cornell University). OpenAlex counts 64 citations as of 19 November 2026. It is open access: the full text is free to read at arXiv (Cornell University).

Citations
64
Most cited in
2025 (14)
References
40
Authors
5

Abstract

This work presents Origami, which provides privacy-preserving inference for large deep neural network (DNN) models through a combination of enclave execution, cryptographic blinding, interspersed with accelerator-based computation. Origami partitions the ML model into multiple partitions. The first partition receives the encrypted user input within an SGX enclave. The enclave decrypts the input and then applies cryptographic blinding to the input data and the model parameters. Cryptographic blinding is a technique that adds noise to obfuscate data. Origami sends the obfuscated data for computation to an untrusted GPU/CPU. The blinding and de-blinding factors are kept private by the SGX enclave, thereby preventing any adversary from denoising the data, when the computation is offloaded to a GPU/CPU. The computed output is returned to the enclave, which decodes the computation on noisy data using the unblinding factors privately stored within SGX. This process may be repeated for each DNN layer, as has been done in prior work Slalom. However, the overhead of blinding and unblinding the data is a limiting factor to scalability. Origami relies on the empirical observation that the feature maps after the first several layers can not be used, even by a powerful conditional GAN adversary to reconstruct input. Hence, Origami dynamically switches to executing the rest of the DNN layers directly on an accelerator without needing any further cryptographic blinding intervention to preserve privacy. We empirically demonstrate that using Origami, a conditional GAN adversary, even with an unlimited inference budget, cannot reconstruct the input. We implement and demonstrate the performance gains of Origami using the VGG-16 and VGG-19 models. Compared to running the entire VGG-19 model within SGX, Origami inference improves the performance of private inference from 11x while using Slalom to 15.1x.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Narra, K. G., Lin, Z., Wang, Y., Balasubramaniam, K., & Annavaram, M. (2019). Privacy-Preserving Inference in Machine Learning Services Using Trusted Execution Environments. arXiv (Cornell University). https://doi.org/10.48550/arxiv.1912.03485

MLA 9

Narra, Krishna Giri, et al. “Privacy-Preserving Inference in Machine Learning Services Using Trusted Execution Environments.” arXiv (Cornell University), 2019. https://doi.org/10.48550/arxiv.1912.03485.

Chicago

Narra, Krishna Giri, Zhifeng Lin, Yongqin Wang, Keshav Balasubramaniam, and Murali Annavaram. 2019. “Privacy-Preserving Inference in Machine Learning Services Using Trusted Execution Environments.” arXiv (Cornell University). https://doi.org/10.48550/arxiv.1912.03485.

Harvard

Narra, K. G. et al. (2019) ‘Privacy-Preserving Inference in Machine Learning Services Using Trusted Execution Environments’, arXiv (Cornell University). doi:10.48550/arxiv.1912.03485.

IEEE

K. G. Narra, Z. Lin, Y. Wang, K. Balasubramaniam, and M. Annavaram, “Privacy-Preserving Inference in Machine Learning Services Using Trusted Execution Environments,” arXiv (Cornell University), 2019, doi: 10.48550/arxiv.1912.03485.

BibTeX

@misc{narra2019privacy,
  title     = {{Privacy-Preserving Inference in Machine Learning Services Using Trusted Execution Environments}},
  author    = {Narra, Krishna Giri and Lin, Zhifeng and Wang, Yongqin and Balasubramaniam, Keshav and Annavaram, Murali},
  howpublished = {arXiv (Cornell University)},
  year      = {2019},
  doi       = {10.48550/arxiv.1912.03485},
  url       = {https://doi.org/10.48550/arxiv.1912.03485}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Citations by year

YearCitationsShare of the best year
202055
202188
202255
20231111
20241313
20251414
202688

Topic and keywords

Physical Sciences › Computer Science › Hardware and Architecture › Security and Verification in Computing

  • privacy-preserving inference
  • trusted execution environments
  • Intel SGX
  • deep neural network
  • model partitioning
  • input reconstruction
  • VGG16
  • VGG19

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Narra, K. G., Lin, Z., Wang, Y., Balasubramaniam, K., & Annavaram, M. (2019). Privacy-Preserving Inference in Machine Learning Services Using Trusted Execution Environments. arXiv (Cornell University). https://doi.org/10.48550/arxiv.1912.03485

How many times has this paper been cited?

64 times, according to OpenAlex on 19 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at arXiv (Cornell University): sign in and open the PDF from this page.

Who wrote this paper?

Krishna Giri Narra, Zhifeng Lin, Yongqin Wang, Keshav Balasubramaniam, Murali Annavaram.

What is the DOI of this paper?

10.48550/arxiv.1912.03485, permanently at https://doi.org/10.48550/arxiv.1912.03485. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 19 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.