Abstract: Efficient acceleration of Deep Neural Networks is a manifold task. In order to save memory requirements and reduce energy consumption we propose the use of dedicated accelerators with novel ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results