deep neural networks (dnns) have advanced in many real-world applications, such as healthcare and autonomous driving. However, their high computational complexity and vulnerability to →