1 paper · 1 filter
Elias Trommer, Bernd Waschneck, Akash Kumar
This work explores the search for heterogeneous approximate multiplier configurations for neural networks that produce high accuracy and low energy consumption. We discuss the vali…