Memristor-based neuromorphic circuit for visual emotional non-associative learning.

Area of Science:

• Neuroscience
• Electrical Engineering
• Artificial Intelligence

Background:

• Current nonassociative learning memristive neural networks often neglect emotional integration.
• Biological characteristics of emotional learning are crucial for advanced AI.

Purpose of the Study:

• To design a nonassociative learning memristive neural network circuit capable of processing visual stimuli to induce emotions.
• To implement functions like habituation, sensitization, and spontaneous recovery within the circuit.

Main Methods:

• A novel circuit comprising voltage control, synapse, potentiation of habituation, and spontaneous recovery modules was designed.
• The circuit was tested using visual stimuli of varying intensities to induce emotions.
• PSpice simulations were employed to validate circuit functionality and accuracy.

Main Results:

• The designed circuit successfully implemented emotional habituation and sensitization in response to visual inputs.
• Simulation results confirmed the circuit's ability to mimic human emotional learning processes.
• The circuit demonstrated functions including habituation, dishabituation, potentiation, and spontaneous recovery.

Conclusions:

• The proposed memristive neural network circuit effectively integrates emotional processing with nonassociative learning.
• This circuit serves as a potential neuromorphic module for robotic affective perception and emotional intelligence.
• The research provides a valuable reference for future studies in brain-inspired emotional AI.