Bioinspired Photoresponsive Single Transistor Neuron for a Neuromorphic Visual System

Abstract

Realizing a neuromorphic-based artificial visual system with low-cost hardware requires a neuromorphic device that can react to light stimuli. This study introduces a photoresponsive neuron device composed of a single transistor, developed by engineering an artificial neuron that responds to light, just like retinal neurons. Neuron firing is activated primarily by electrical stimuli such as current via a well-known single transistor latch phenomenon. Its firing characteristics, represented by spiking frequency and amplitude, are additionally modulated by optical stimuli such as photons. When light is illuminated onto the neuron transistor, electron−hole pairs are generated, and they allow the neuron transistor to fire at lower firing threshold voltage. Different photoresponsive properties can be modulated by the intensity and wavelength of the light, analogous to the behavior of retinal neurons. The artificial visual system can be miniaturized because a photoresponsive neuronal function is realized without bulky components such as image sensors and extra circuits.

Retina‐Inspired Self‐Powered Artificial Optoelectronic Synapses with Selective Detection in Organic Asymmetric Heterojunctions

“…In the vertebrate retina, abundant sensory neurons connected by synapses are distributed within different layers, mainly including photoreceptors, bipolar cell layers, and ganglion cell layers ( Figure 1 a ). [ 14 ] Photoreceptors such as rods and cones convert incoming light into electrical signals, especially for cone cells, which can detect light of specific wavelengths. [ 1 ] Then, the signals are transferred into the bipolar cells by electronic‐to‐ionic conversion.…”

“…[ 8 ] Therefore, with the increasing demand for edge computing in the era of big data, various retina‐inspired neuromorphic devices have been proposed to provide a promising pathway toward artificial visual systems with high‐efficiency signal processing. [ 12 , 13 , 14 , 15 , 16 ] Recently, optoelectronic synaptic devices with spectral selectivity and multispectral sensing capabilities have attracted considerable attention because they can combine the dual functions of sensing and preprocessing in a single device. [ 1 , 17 , 18 , 19 , 20 ] Note that the sustainable retina‐imitating intelligent operation of these devices to cope with ubiquitous sensing still requires an external electrical supply for energy harvesting, energy conversion, and information transmission to obtain useful visual signals.…”

Retinomorphic optoelectronic devices for intelligent machine vision

“…As shown in Figure 1 , the human visual system mainly consists of the eyeballs, transmission nerve, and visual cortex of the brain. Light from the environment and external objects enters the crystalline lens through the pupil at the front of the eyeball, and finally reaches the retina after refraction ( Han et al., 2020 ). In particular, the retina has a clear hierarchical structure for photoelectric information conversion, preprocessing, and transfer.…”

“…This fine retinal cell hierarchical structure gives the human eye a variety of functions such as perception, signal classification and integration, preprocessing, etc.

Figure 1 Schematic illustration of the biological visual perception system Reproduced with permission ( Han et al., 2020 ). Copyright 2020, American Chemical Society.

…”

A Single Transistor‐Based Threshold Switch for a Bio‐Inspired Reconfigurable Threshold Logic

“…On the other hand, a single transistor‐based neuron with a property of leaky integrate‐and‐fire (LIF) was recently demonstrated with various structures. [ 18–21 ] The phenomenon of single transistor latch (STL) occurring in a single metal‐oxide‐semiconductor field‐effect transistor (MOSFET) with a floating body can enable threshold switching, and additional circuitry such as comparator or Schmitt trigger is thus not needed. In this work, bio‐inspired reconfigurable threshold logic operations are demonstrated using a single transistor‐based threshold switch abbreviated here as “threshold switch.” Linear Boolean logic functions including AND, OR, NOT, NAND, and NOR are experimentally implemented.…”