# PrivCirNet: Efficient Private Inference via Block Circulant Transformation

Source: [PrivCirNet: Efficient Private Inference via Block Circulant Transformation](<https://www.sota2.com/research/paper/68fa756a1e209131cea67ba2>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T03:33:53.801Z

Authors: Tianshi Xu, Lemeng Wu, Runsheng Wang, Meng Li

Published: 2024-05-23T13:44:48.000Z

## Abstract

Homomorphic encryption (HE)-based deep neural network (DNN) inference protects data and model privacy but suffers from significant computation overhead. We observe transforming the DNN weights into circulant matrices converts general matrix-vector multiplications into HE-friendly 1-dimensional convolutions, drastically reducing the HE computation cost. Hence, in this paper, we propose \\method, a protocol/network co-optimization framework based on block circulant transformation. At the protocol level, PrivCirNet customizes the HE encoding algorithm that is fully compatible with the block circulant transformation and reduces the computation latency in proportion to the block size. At the network level, we propose a latency-aware formulation to search for the layer-wise block size assignment based on second-order information. PrivCirNet also leverages layer fusion to further reduce the inference cost. We compare PrivCirNet with the state-of-the-art HE-based framework Bolt (IEEE S\\&P 2024) and the HE-friendly pruning method SpENCNN (ICML 2023). For ResNet-18 and Vision Transformer (ViT) on Tiny ImageNet, PrivCirNet reduces latency by $5.0\\times$ and $1.3\\times$ with iso-accuracy over Bolt, respectively, and improves accuracy by $4.1\\%$ and $12\\%$ over SpENCNN, respectively. For MobileNetV2 on ImageNet, PrivCirNet achieves $1.7\\times$ lower latency and $4.2\\%$ better accuracy over Bolt and SpENCNN, respectively. Our code and checkpoints are available on Git Hub.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2405.14569>)

[PDF](<https://arxiv.org/pdf/2405.14569.pdf>)