.png){kind=icon}
.png){kind=icon}
Return
Exploiting edge features for transferable adversarial attacks in distributed machine learning
.png){kind=icon}
PRE
AI
DOI：10.1016/j.iot.2025.101795.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Novel threat model for black-box attacks against edge-cloud partitioned framework. • Lightweight shape reconstruction of vectorized communication-channel data-tensors. • Edge-cloud partitioned framework stealing via feature-based surrogate distillation. • Methodological advances of black-box transfer attacks via internal features knowledge.