APMSA: Adversarial perturbation against model stealing attacks
Training a Deep Learning (DL) model requires proprietary data and computing-intensive
resources. To recoup their training costs, a model provider can monetize DL models through …
resources. To recoup their training costs, a model provider can monetize DL models through …
SeInspect: Defending model stealing via heterogeneous semantic inspection
X Liu, Z Ma, Y Liu, Z Qin, J Zhang, Z Wang - European Symposium on …, 2022 - Springer
Recent works developed an emerging attack, called Model Stealing (MS), to steal the
functionalities of remote models, rendering the privacy of cloud-based machine learning …
functionalities of remote models, rendering the privacy of cloud-based machine learning …
Model Stealing Detection for IoT Services Based on Multi-Dimensional Features
X Liu, T Liu, H Yang, J Dong, Z Ying… - IEEE Internet of Things …, 2024 - ieeexplore.ieee.org
Model stealing (MS) attacks pose a significant security concern for machine learning models
on cloud platforms, as they can reconstruct a substitute model with limited effort to evade …
on cloud platforms, as they can reconstruct a substitute model with limited effort to evade …
Adaptive and robust watermark against model extraction attack
K Pang, T Qi, C Wu, M Bai - arXiv preprint arXiv:2405.02365, 2024 - arxiv.org
Large language models have boosted Large Models as a Service (LMaaS) into a thriving
business sector. But even model owners offering only API access while keeping model …
business sector. But even model owners offering only API access while keeping model …