A Bionic Circuit for Emotion Associative Memory Based on Memristive Neural Networks

Bionics-driven brain-inspired intelligence is a vital component of the field of artificial intelligence. Constructing bionic circuits for emotional associative memory enables the simulation of human brain functions such as learning, forgetting, and memory. Two similar emotions, namely fear and disgust, are introduced in this study. Since the synaptic structures between different emotions vary, distinct emotional modules are designed to simulate these two similar emotions. The circuit consists of four modules: the fear emotion logic module, which selects corresponding functions by determining the activation states of fear-related language and auditory signals; the fear emotion synapse module, responsible for implementing the learning or forgetting functions of fear; the disgust emotion logic module, which selects corresponding functions by determining the activation states of disgust-related language and auditory signals; and the disgust emotion synapse module, responsible for implementing the learning or forgetting functions of disgust. The effects of these two similar emotions on associative memory were evaluated using PSpice simulations, demonstrating distinct impacts on emotional associative memory.