Artificial neural network using thin-film transistors - Working confirmation of asymmetric circuit -

Abstract

We are developing neural networks of device level using thin-film transistors (TFT). By adopting an interconnect-type neural network and utilizing a characteristic shift of poly-Si TFTs as a variable strength of synapse connection, which was originally an issue, we realized the neuron consisting of eight TFTs and synapse of only one TFT. Particularly in this presentation, we confirmed the working by a circuit where the input and output elements are asymmetric. This is a result leading to a super-large, self-learning, and high-flexibility system.