Caiqin Dong (Jinan University), Jian Weng (Jinan University), Jia-Nan Liu (Jinan University), Yue Zhang (Jinan University), Yao Tong (Guangzhou Fongwell Data Limited Company), Anjia Yang (Jinan University), Yudan Cheng (Jinan University), Shun Hu (Jinan University)

In secure machine learning inference, most of the schemes assume that the server is semi-honest (honestly following the protocol but attempting to infer additional information). However, the server may be malicious (e.g., using a low-quality model or deviating from the protocol) in the real world. Although a few studies have considered a malicious server that deviates from the protocol, they ignore the verification of model accuracy (where the malicious server uses a low-quality model) meanwhile preserving the privacy of both the server's model and the client's inputs. To address these issues, we propose textit{Fusion}, where the client mixes the public samples (which have known query results) with their own samples to be queried as the inputs of multi-party computation to jointly perform the secure inference. Since a server that uses a low-quality model or deviates from the protocol can only produce results that can be easily identified by the client, textit{Fusion} forces the server to behave honestly, thereby addressing all those aforementioned issues without leveraging expensive cryptographic techniques. Our evaluation indicates that textit{Fusion} is 48.06$times$ faster and uses 30.90$times$ less communication than the existing maliciously secure inference protocol (which currently does not support the verification of the model accuracy). In addition, to show the scalability, we conduct ImageNet-scale inference on the practical ResNet50 model and it costs 8.678 minutes and 10.117 GiB of communication in a WAN setting, which is 1.18$times$ faster and has 2.64$times$ less communication than those of the semi-honest protocol.

View More Papers

The Walls Have Ears: Gauging Security Awareness in a...

Gokul Jayakrishnan, Vijayanand Banahatti, Sachin Lodha (TCS Research Tata Consultancy Services Ltd.)

Read More

Assessing the Impact of Interface Vulnerabilities in Compartmentalized Software

Hugo Lefeuvre (The University of Manchester), Vlad-Andrei Bădoiu (University Politehnica of Bucharest), Yi Chen (Rice University), Felipe Huici (Unikraft.io), Nathan Dautenhahn (Rice University), Pierre Olivier (The University of Manchester)

Read More

VulHawk: Cross-architecture Vulnerability Detection with Entropy-based Binary Code Search

Zhenhao Luo (College of Computer, National University of Defense Technology), Pengfei Wang (College of Computer, National University of Defense Technology), Baosheng Wang (College of Computer, National University of Defense Technology), Yong Tang (College of Computer, National University of Defense Technology), Wei Xie (College of Computer, National University of Defense Technology), Xu Zhou (College of Computer,…

Read More