Fig. 6: Performance of the trained FE-PS-NET.
From: In situ training of an in-sensor artificial neural network based on ferroelectric photosensors
a Photographs of the vehicle when executing different motions (top) along with the corresponding Iouts of the FE-PS-NET (bottom). b Long-term stability of the Iouts of the FE-PS-NET. c Recognition accuracies of the FE-PS-NET at different noise levels (noise level refers to the standard deviation of the Gaussian noise). d Schematic architecture of the FE-PS-NET (left) and its inference speed (right). e Schematic architecture of a von Neumann machine vision system (left) and its inference speed (right).
