[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119778-en":3,"doc-seo-119778-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},119778,962075006959,"Anda","https://ap-avatar.wpscdn.com/avatar/e0002397efbe92a78e?_k=1776741047341049297",8,"Research & Report","Efficient Privacy-Preserving Machine Learning with Lightweight Trusted Hardware","A secure machine learning inference platform is proposed using a small dedicated security processor, enabling easier protection and deployment than today’s TEEs in high-performance CPUs. The platform delivers substantial speed and communication improvements over leading distributed PPML protocols, achieves security via abort against malicious adversaries under honest majority, and removes reliance on TEE secure-memory size. The approach scales to modern neural networks such as ResNet18 and Transformers.","Efficient Privacy-Preserving Machine Learning with Lightweight Trusted Hardware  \nPengzhi Huang Cornell University [ph448@cornell.edu](ph448@cornell.edu)  \nThang Hoang  \nVirginia Tech [thanghoang@vt.edu](thanghoang@vt.edu)  \nYueying Li Cornell University [yl3469@cornell.edu](yl3469@cornell.edu)  \nElaine Shi  \nCarnegie Mellon University [runting@gmail.com](runting@gmail.com)  \nG. Edward Suh  \nCornell University / Meta AI [edsuh@meta.com](edsuh@meta.com)  \narXiv :2210 . 10 133v 3 [ cs .CR] 8 Sep 2023  \nAbstract—In this paper, we propose a new secure machine learning inference platform assisted by a small dedicated security processor, which will be easier to protect and deploy compared to today’s TEEs integrated into high-performance processors. Our platform provides three main advantages over the state-of-the-art: (i) We achieve significant performance improvements compared to state-of-the-art distributed PrivacyPreserving Machine Learning (PPML) protocols, with only a small security processor that is comparable to a discrete security chip such as the Trusted Platform Module (TPM) or on-chip security subsystems in SoCs similar to the Apple enclave processor. In the semi-honest setting with WAN/GPU, our scheme is 4×-63× faster than Falcon (PoPETs’21) and AriaNN (PoPETs’22) and 3.8×-12× more communication efficient. We achieve even higher performance improvements in the malicious setting. (ii) Our platform guarantees security with abort against malicious adversaries under honest majority assumption. (iii) Our technique is not limited by the size of secure memory in a TEE and can support high-capacity modern neural networks like ResNet18 and Transformer. While previous work investigated the use of high-performance TEEs in PPML, this work represents the first to show that even tiny secure hardware with really limited performance can be leveraged to significantly speed-up distributed PPML protocols if the protocol can be carefully designed for lightweight trusted hardware.  \n1. Introduction  \nAs the world increasingly relies on machine learning for everyday tasks, a large amount of potentially sensitive or private data need to be processed by machine learning algorithms. For example, machine learning models for medical applications may need to use private datasets distributed in multiple nations as inputs [37] . A cloud-based machine learning services process private data from users with pretrained models to provide predictions [1, 22] . The data to be shared in these applications are often private and sensitive and must be protected from the risk of leakage. Government regulations may play an essential role as a policy, but cannot guarantee actual protection. We need technical protection for privacy-preserving machine learning (PPML) for strong confidentiality and privacy guarantees.  \nIn this paper, we propose a new PPML framework, named STAMP (Small Trusted hardware Assisted MPc), which enables far more efficient secure multiparty computation (MPC) for machine learning through a novel use of small lightweight trusted hardware (LTH) . MPC refers to a protocol that allows multiple participants to jointly evaluate a particular problem while keeping their inputs from being revealed to each other. Ever since Yao’s initial studies (later called Garbled Circuit) [89, 90] which gave such a secure protocol in the case of two semi-honest parties, many studies have been conducted to improve the efficiency, to expand to more than two parties, and to ensure the feasibility against malicious behaviors. Recently, there has been significant interest in using and optimizing MPC for secure machine learning computation [54, 82, 83, 67, 40] . However, the overhead for MPC-based PPML is still significant.  \nFor low-overhead secure computation, the trusted execution environments (TEEs) in modern microprocessors such as Intel SGX [12] AMD SEV [68] aim to provide hardwarebased protection for the confidentiality and integrity of data and code inside. If t","cbCailNUfxkqCXw3","https://ap.wps.com/l/cbCailNUfxkqCXw3","pdf",831630,1,22,"English","en",105,"# Introduction\n## Motivation for privacy-preserving machine learning\n## PPML via MPC and overhead challenges\n## Trusted execution environments and their limitations\n## Lightweight trusted hardware and STAMP framework","[{\"question\":\"What is the main contribution of this paper’s proposed platform?\",\"answer\":\"It presents a secure machine learning inference platform that leverages a small dedicated security processor to reduce MPC overhead compared with using high-performance TEEs.\"},{\"question\":\"How does the proposed approach improve performance and communication efficiency?\",\"answer\":\"It achieves large speedups over state-of-the-art distributed PPML protocols and improves communication efficiency, with stronger gains also reported in the malicious setting.\"},{\"question\":\"What security guarantees does the platform provide?\",\"answer\":\"Under an honest majority assumption, it guarantees security with abort against malicious adversaries.\"}]","Efficient Privacy-Preserving Machine Learning with Lightweight Trusted Hardware | PDF",1785726262,55,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"efficient-privacy-preserving-machine-learning-with-lightweight-trusted-hardware","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/efficient-privacy-preserving-machine-learning-with-lightweight-trusted-hardware/119778/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is the main contribution of this paper’s proposed platform?","Question",{"text":75,"@type":76},"It presents a secure machine learning inference platform that leverages a small dedicated security processor to reduce MPC overhead compared with using high-performance TEEs.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the proposed approach improve performance and communication efficiency?",{"text":80,"@type":76},"It achieves large speedups over state-of-the-art distributed PPML protocols and improves communication efficiency, with stronger gains also reported in the malicious setting.",{"name":82,"@type":73,"acceptedAnswer":83},"What security guarantees does the platform provide?",{"text":84,"@type":76},"Under an honest majority assumption, it guarantees security with abort against malicious adversaries.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]