A novel memristor-based bionic neural network circuit with crossmodal integration and forgetting effects.

Area of Science:

• Artificial Intelligence
• Neuroscience
• Electrical Engineering

Background:

• Brain cognitive functions rely on cross-modal interactions between learning and memory.
• Visual-olfactory associative memory provides a model for understanding these brain mechanisms.

Purpose of the Study:

• To propose a cognition-driven bionic cross-modal associative memristor neural network (CAMNN).
• To simulate the complete processes of learning, memory, and forgetting in the brain.
• To model cross-modal associative memory interactions.

Main Methods:

• Designed a circuit with voltage control, synaptic (memristor-based), and delay modules.
• Implemented sensory information input, integration, and long-term response maintenance.
• Utilized memristors to simulate synaptic weights for associative memory.

Main Results:

• The CAMNN achieved functions including threshold detection, unimodal learning, cross-modal learning, and cross-modal reinforcement learning.
• PSPICE simulations confirmed the circuit's ability to simulate cross-modal interactions.
• Demonstrated the network's capacity to mimic brain memory processes.

Conclusions:

• The proposed CAMNN effectively simulates brain-inspired cross-modal associative memory.
• This work offers a reference for developing intelligent robots and neural network circuits.
• Highlights the potential of memristor-based networks for AI.