paper-with-me

홈 › Papers

Privatar: Scalable Privacy-preserving Multi-user VR via Secure Offloading

2026-04-19 · Jianming Tong, Hanshen Xiao, Krishna Kumar Nair, Hao Kang, Ashish Sirasao, Ziqi Zhang, G. Edward Suh, Tushar Krishna arxiv

Multi-user virtual reality enables immersive interaction. However, rendering avatars for numerous participants on each headset incurs prohibitive computational overhead, limiting scalability. We introduce a framework, Privatar, to offload avatar reconstruction from headset to untrusted devices within the same local network while safeguarding attacks against adversaries capable of intercepting offloaded data. Privatar's key insight is that domain-specific knowledge of avatar reconstruction enables provably private offloading at minimal cost. (1) System level. We observe avatar reconstruction is frequency-domain decomposable via BDCT with negligible quality drop, and propose Horizontal Partitioning (HP) to keep high-energy frequency components on-device and offloads only low-energy components. HP offloads local computation while reducing information leakage to low-energy subsets only. (2) Privacy level. For individually offloaded, multi-dimensional signals without aggregation, worst-case local Differential Privacy requires prohibitive noise, ruining utility. We observe users' expression statistical distribution are slowly changing over time and trackable online, and hence propose Distribution-Aware Minimal Perturbation. DAMP minimizes noise based on each user's expression distribution to significantly reduce its effects on utility, retaining formal privacy guarantee. Combined, HP provides empirical privacy against expression identification attacks. DAMP further augments it to offer a formal guarantee against arbitrary adversaries. On a Meta Quest Pro, Privatar supports 2.37x more concurrent users at 6.5% higher reconstruction loss and 9% energy overhead, providing a better throughout-loss Pareto frontier over quantization, sparsity and local construction baselines. Privatar provides both provable privacy guarantee and stays robust against both empirical and NN-based attacks.

📄 PDF Abstract BibTeX arXiv:2604.17476

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

SecureBoost: A Lossless Federated Learning Framework

2019-01-25 · Kewei Cheng, Tao Fan, Yilun Jin, Yang Liu 외

The protection of user privacy is an important concern in machine learning, as evidenced by the rolling out of the General Data Protection Regulation (GDPR) in the European Union (EU) in May 2018. The GDPR is designed to…

BIG-bench Machine LearningEntity AlignmentFederated LearningPrivacy Preserving

zPROBE: Zero Peek Robustness Checks for Federated Learning

2022-06-24 · ICCV 2023 1 · Zahra Ghodsi, Mojan Javaheripi, Nojan Sheybani, Xinqiao Zhang 외

Privacy-preserving federated learning allows multiple users to jointly train a model with coordination of a central server. The server only learns the final aggregation result, thus the users' (private) training data is …

Federated LearningPrivacy Preserving

An Efficient Privacy-Preserving Multi-Keyword Query Scheme in Location Based Services

2020-08-21 · IEEE 2020 8 · SHIWEN ZHANG 1, 2, (Member, TINGTING YAO3 외

With the proliferation of location-aware mobile devices and the prevalence of wireless communications, location-based services (LBS) have attracted much particular attention in recent years. For flexibility and cost sa…

Privacy Preserving

LSRP: A Leader-Subordinate Retrieval Framework for Privacy-Preserving Cloud-Device Collaboration

2025-05-08 · Yingyi Zhang, Pengyue Jia, Xianneng Li, Derong Xu 외

Cloud-device collaboration leverages on-cloud Large Language Models (LLMs) for handling public user queries and on-device Small Language Models (SLMs) for processing private user data, collectively forming a powerful and…

Privacy PreservingRAGRetrievalRetrieval-augmented Generation

Federated Learning of User Authentication Models

2020-07-09 · Hossein Hosseini, Sungrack Yun, Hyunsin Park, Christos Louizos 외

Machine learning-based User Authentication (UA) models have been widely deployed in smart devices. UA models are trained to map input data of different users to highly separable embedding vectors, which are then used to …

Federated LearningPrivacy PreservingSpeaker Verification