no code implementations • 19 Oct 2018 • Yuriy Mishchenko, Yusuf Goren, Ming Sun, Chris Beauchene, Spyros Matsoukas, Oleg Rybakov, Shiv Naga Prasad Vitaladevuni
We investigate low-bit quantization to reduce computational cost of deep neural network (DNN) based keyword spotting (KWS).