Publications

A unified approximation framework for compressing and accelerating deep neural networks