Impact of Low-bitwidth Quantization on the Adversarial Robustness for Embedded Neural Networks

Open in new window