Abstract: This research introduces a novel Processing Engine (PE) for convolution layers in Deep Neural Network (DNN) accelerators. The PE integrates multi-precision dynamic fraction fixed-point ...
Abstract: Using time-difference-of-arrival (TDOA) measurements from several sensors to locate a target is one of the most widely used wireless localization methods. A challenge for TDOA localization ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results