Attention-driven Object Recognition

Project Overview

This is a computational neuroscience project explores the role of attention mechanisms in object recognition, drawing inspiration from biological systems to enhance AI models. By integrating spatial attention pathways into an object recognition model with additional visual cortex layers (V1-V4, Spat1, Spat2), we demonstrate that attention improves accuracy in complex tasks, though often with increased reaction times.