Bioinspired Adaptive Neuron Enabled by Self-powered Optoelectronic Memristor and Threshold Switching Memory for...

Area of Science:

• Materials Science
• Neuroscience
• Electronics

Background:

• Visual adaptation is crucial for organisms to perceive accurately in diverse light conditions.
• Bioinspired electronics with adaptive capabilities are needed to replicate visual system functions.

Purpose of the Study:

• To develop a self-powered optoelectronic memristor exhibiting visual adaptation.
• To construct a bioinspired visual adaptive neuron for processing light stimuli.
• To demonstrate adaptive image preprocessing for enhanced object recognition.

Main Methods:

• Fabrication of a ZnO/WOx heterojunction optoelectronic memristor.
• Utilizing photovoltaic effects and electron trapping for adaptive functions.
• Integration with a NbOx-based threshold switching memory to create a visual adaptive neuron.

Main Results:

• The developed memristor demonstrated visual adaptation functions, including desensitization and Weber's law.
• A bioinspired visual adaptive neuron successfully converted light stimuli into dynamic spike trains.
• Adaptive image preprocessing improved object recognition accuracy for overexposed images.

Conclusions:

• The study offers a novel approach for creating biologically plausible visual adaptation in electronic systems.
• This work advances the development of dynamic neuromorphic visual systems.
• The developed device shows potential for future applications in intelligent vision systems.