3 papers
cs.CR2024
A Semi Black-Box Adversarial Bit-Flip Attack with Limited DNN Model Information
Behnam Ghavami, Mani Sadati, Mohammad Shahidzadeh +2
Despite the rising prevalence of deep neural networks (DNNs) in cyber-physical systems, their vulnerability to adversarial bit-flip attacks (BFAs) is a noteworthy concern. This pap…
cs.AI2024
ZOBNN: Zero-Overhead Dependable Design of Binary Neural Networks with Deliberately Quantized Parameters
Behnam Ghavami, Mohammad Shahidzadeh, Lesley Shannon +1
Low-precision weights and activations in deep neural networks (DNNs) outperform their full-precision counterparts in terms of hardware efficiency. When implemented with low-precisi…
cs.LG2024
DNN Memory Footprint Reduction via Post-Training Intra-Layer Multi-Precision Quantization
Behnam Ghavami, Amin Kamjoo, Lesley Shannon +1
The imperative to deploy Deep Neural Network (DNN) models on resource-constrained edge devices, spurred by privacy concerns, has become increasingly apparent. To facilitate the tra…