International Conference on Computer and Knowledge Engineering

Home / 15th International Conference on Computer and Knowledge Engineering

Simulating Human Visual Cortex and Recall System with Convolutional Neural Networks

Authors :

Sina Saadati¹ Abdolah Sepahvand²

1- Amirkabir University of Technology (Tehran Polytechnic) 2- Amirkabir University of Technology (Tehran Polytechnic)

Keywords :

Computational psychology،Neural networks in computational cognitive،Artificial neural networks،Human brain simulation

Abstract :

Artificial neural networks (ANNs) are mathematical representations inspired by biological neurons, the foundational units of the human and animal brain. While ANNs are extensively applied in tasks such as image processing, classification, semantic segmentation, localization, and image generation, they also offer potential for simulating psychological and cognitive processes. In this research, we examine whether convolutional neural networks (CNNs), commonly used in image processing tasks, can mimic the visual recall mechanism of the brain. A deep CNN was trained to associate a series of input images with a specific output image, referred to as the “memory image.” After training, we fed the network new input images and observed the generated outputs. In repeated trials, the CNN consistently produced images that were visually similar to the last memory image. These results suggest that CNNs may offer a viable model for exploring how the brain recalls and reconstructs recent visual memories. Consequently, the artificial neural networks have the potential to simulate psychological behaviors caused by the brain.