Memristor-Based Neural Network Circuit With Flashbulb Memory Effect

Most existing memristor-based neural network designs have primarily focused on how different types of emotions influence memory recall speed. However, the role of emotional intensity in memory processes remains insufficiently explored. In this study, we propose a dual-channel brain-inspired neural network circuit based on memristors, which incorporates both the flashbulb memory effect and emotion-enhanced mechanisms. The proposed design not only considers the impact of various emotional types across different intensity levels, but also integrates the modulatory effects of linguistic factors on memory performance. The circuit comprises three functionally distinct modules: an emotion module, an emotion enhancement module, and a language enhancement module. Specifically: The emotion module is responsible for producing different emotional states; The emotion enhancement module enables memory reinforcement through emotional intensity and triggers the flashbulb memory effect; The language enhancement module improves recall speed by introducing diverse language inputs. This design can be broadly applied to the development of emotionally responsive biomimetic robots and offers valuable insights for advancing brain-inspired memory systems.