FFConv: Fast Factorized Neural Network Inference on Encrypted Data
- URL: http://arxiv.org/abs/2102.03494v1
- Date: Sat, 6 Feb 2021 03:10:13 GMT
- Title: FFConv: Fast Factorized Neural Network Inference on Encrypted Data
- Authors: Yuxiao Lu, Jie Lin, Chao Jin, Zhe Wang, Khin Mi Mi Aung, Xiaoli Li
- Abstract summary: We propose a low-rank factorization method called FFConv to unify convolution and ciphertext packing.
Compared to prior art LoLa and Falcon, our method reduces the inference latency by up to 87% and 12%, respectively.
- Score: 9.868787266501036
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: Homomorphic Encryption (HE), allowing computations on encrypted data
(ciphertext) without decrypting it first, enables secure but prohibitively slow
Neural Network (HENN) inference for privacy-preserving applications in clouds.
To reduce HENN inference latency, one approach is to pack multiple messages
into a single ciphertext in order to reduce the number of ciphertexts and
support massive parallelism of Homomorphic Multiply-Add (HMA) operations
between ciphertexts. However, different ciphertext packing schemes have to be
designed for different convolution layers and each of them introduces overheads
that are far more expensive than HMA operations. In this paper, we propose a
low-rank factorization method called FFConv to unify convolution and ciphertext
packing. To our knowledge, FFConv is the first work that is capable of
accelerating the overheads induced by different ciphertext packing schemes
simultaneously, without incurring a significant increase in noise budget.
Compared to prior art LoLa and Falcon, our method reduces the inference latency
by up to 87% and 12%, respectively, with comparable accuracy on MNIST and
CIFAR-10.
Related papers
- Semi-Compressed CRYSTALS-Kyber [4.317605401561789]
We show it is feasible to reduce the communication overhead of the Kyber by 54%.
The improvement is based on two technologies: ciphertext quantization and plaintext encoding.
We show that with the Lloyd-Max quantization, 8-PAM, Gray mapping, and a shortened binary BCH(768,638,13) code, the proposed scheme encapsulates 638 bits in a single ciphertext.
arXiv Detail & Related papers (2024-07-25T00:54:22Z) - Parallel Decoding via Hidden Transfer for Lossless Large Language Model Acceleration [54.897493351694195]
We propose a novel parallel decoding approach, namely textithidden transfer, which decodes multiple successive tokens simultaneously in a single forward pass.
In terms of acceleration metrics, we outperform all the single-model acceleration techniques, including Medusa and Self-Speculative decoding.
arXiv Detail & Related papers (2024-04-18T09:17:06Z) - Coding-Based Hybrid Post-Quantum Cryptosystem for Non-Uniform Information [53.85237314348328]
We introduce for non-uniform messages a novel hybrid universal network coding cryptosystem (NU-HUNCC)
We show that NU-HUNCC is information-theoretic individually secured against an eavesdropper with access to any subset of the links.
arXiv Detail & Related papers (2024-02-13T12:12:39Z) - SOCI^+: An Enhanced Toolkit for Secure OutsourcedComputation on Integers [50.608828039206365]
We propose SOCI+ which significantly improves the performance of SOCI.
SOCI+ employs a novel (2, 2)-threshold Paillier cryptosystem with fast encryption and decryption as its cryptographic primitive.
Compared with SOCI, our experimental evaluation shows that SOCI+ is up to 5.4 times more efficient in computation and 40% less in communication overhead.
arXiv Detail & Related papers (2023-09-27T05:19:32Z) - Encrypted Dynamic Control exploiting Limited Number of Multiplications and a Method using Ring-LWE based Cryptosystem [0.3749861135832073]
We present a method to encrypt dynamic controllers that can be implemented through most homomorphic encryption schemes.
We propose a customization of the method for Ring-Learning With Errors (Ring-LWE) based cryptosystems.
arXiv Detail & Related papers (2023-07-07T08:24:48Z) - CipherSniffer: Classifying Cipher Types [0.0]
We frame the decryption task as a classification problem.
We first create a dataset of transpositions, substitutions, text reversals, word reversals, sentence shifts, and unencrypted text.
arXiv Detail & Related papers (2023-06-13T20:18:24Z) - HE-PEx: Efficient Machine Learning under Homomorphic Encryption using
Pruning, Permutation and Expansion [4.209035833239216]
Homomorphic encryption (HE) is a method of performing computations over encrypted data.
We propose a novel set of pruning methods that reduce the latency and memory requirement, thus bringing the effectiveness of pruning methods to HE.
We demonstrate the advantage of our method on fully connected layers where the weights are packed using a recently proposed packing technique called tile tensors.
arXiv Detail & Related papers (2022-07-07T15:49:24Z) - THE-X: Privacy-Preserving Transformer Inference with Homomorphic
Encryption [112.02441503951297]
Privacy-preserving inference of transformer models is on the demand of cloud service users.
We introduce $textitTHE-X$, an approximation approach for transformers, which enables privacy-preserving inference of pre-trained models.
arXiv Detail & Related papers (2022-06-01T03:49:18Z) - Recovering AES Keys with a Deep Cold Boot Attack [91.22679787578438]
Cold boot attacks inspect the corrupted random access memory soon after the power has been shut down.
In this work, we combine a novel cryptographic variant of a deep error correcting code technique with a modified SAT solver scheme to apply the attack on AES keys.
Our results show that our methods outperform the state of the art attack methods by a very large margin.
arXiv Detail & Related papers (2021-06-09T07:57:01Z) - Faster Secure Data Mining via Distributed Homomorphic Encryption [108.77460689459247]
Homomorphic Encryption (HE) is receiving more and more attention recently for its capability to do computations over the encrypted field.
We propose a novel general distributed HE-based data mining framework towards one step of solving the scaling problem.
We verify the efficiency and effectiveness of our new framework by testing over various data mining algorithms and benchmark data-sets.
arXiv Detail & Related papers (2020-06-17T18:14:30Z) - TEDL: A Text Encryption Method Based on Deep Learning [10.428079716944463]
This paper proposes a novel text encryption method based on deep learning called TEDL.
Results of experiments and relevant analyses show that TEDL performs well for security, efficiency, generality, and has a lower demand for the frequency of key redistribution.
arXiv Detail & Related papers (2020-03-09T11:04:36Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.