Publication

Trainable Preprocessing for Reduced Precision Neural Networks