Abstract:Perceptual expertise and attention are two important factors that enable superior object recognition and task performance. While expertise enhances knowledge and provides a holistic understanding of the environment, attention allows us to selectively focus on task-related information and suppress distraction. It has been suggested that attention operates differently in experts and in novices and much remains unknown. This study investigates the relationship between perceptual expertise and attention using convolutional neural networks (CNNs) which are shown to be good models of primate visual pathways. Two CNN models were trained to become experts in either face or scene recognition, and the effect of attention on performance was evaluated in tasks involving complex stimuli, such as merged images containing superimposed faces and scenes. The goal was to explore how feature-based attention (FBA) influences recognition within and outside the models' domain of expertise. Results showed that each model performed better in its area of expertise and that FBA further enhanced task performance only within the domain of expertise, increasing performance by up to 35% in scene recognition and 15% in face recognition. However, attention had reduced or negative effects when applied outside the model's expertise domain. Neural unit-level analysis revealed that expertise leads to stronger neuronal tuning towards category-specific features and sharper tuning curves, reflected in greater representational dissimilarity between targets and distractors, which, according to the biased competition model of attention, reduces competition and leads to enhanced performance. These findings highlight the critical role of neural tuning at the single neuron level as well as at the level of multivariate neural representations in distinguishing the effects of attention in experts and in novices and demonstrate that CNNs can be used fruitfully as computational models for addressing neuroscience questions not practical with the empirical methods.

Emergent human-like covert attention in feedforward convolutional neural networks

Bridging Neurons and Behavior in a Convolutional Neural Network with Emergent Human-like Covert Attention

Deep Neural Networks Evolve Human-like Attention Distribution during Goal-directed Reading Comprehension

Deep Neural Networks Evolve Human-like Attention Distribution during Reading Comprehension.

Guiding Visual Attention in Deep Convolutional Neural Networks Based on Human Eye Movements

Do humans and Convolutional Neural Networks attend to similar areas during scene classification: Effects of task and image type

Perceptual Expertise and Attention: An Exploration using Deep Neural Networks

Human Vs Machine Attention in Neural Networks: A Comparative Study.

Human Eyes Inspired Recurrent Neural Networks are More Robust Against Adversarial Noises

Human Eyes-Inspired Recurrent Neural Networks Are More Robust Against Adversarial Noises

A feedforward mechanism for human-like contour integration

Seeing eye-to-eye? A comparison of object recognition performance in humans and deep convolutional neural networks under image manipulation

Modelling attention control using a convolutional neural network designed after the ventral visual pathway

An Overview of the Attention Mechanisms in Computer Vision

Understanding More about Human and Machine Attention in Deep Neural Networks

Visual Attention driven by Convolutional Features

The Costs and Benefits of Goal-Directed Attention in Deep Convolutional Neural Networks

Focal vs Diffuse: Mechanisms of attention mediated performance enhancement in a hierarchical model of the visual system

Learning to attend in a brain-inspired deep neural network

Passive Attention in Artificial Neural Networks Predicts Human Visual Selectivity

A Spatial–Channel–Temporal-Fused Attention for Spiking Neural Networks