Abstract: Many studies have demonstrated the vulnerability of Deep Neural Networks (DNNs) to adversarial attacks. While numerous research efforts have proposed high-performance adversarial attacks and ...
Abstract: We present a method for implementing high speed finite impulse response (FIR) filters using just registered adders and hardwired shifts. We extensively use a modified common subexpression ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results