Return to previous page

Approximate Softmax Functions for Energy-Efficient Deep Neural Networks