Learning visual biases from human imagination

TL;DR

This paper explores extracting and transferring human visual biases into machine recognition systems, demonstrating improved generalization and performance with limited training data through novel psychophysics-inspired methods.

Contribution

Introduces a new method to estimate and transfer human visual biases into machine classifiers, enhancing recognition performance and generalization.

Findings

Classifiers from human visual biases can be transferred to machines with some success.

Bias-constrained SVMs improve recognition when training data is scarce.

Transferring human biases helps in cross-dataset generalization.

Abstract

Although the human visual system can recognize many concepts under challenging conditions, it still has some biases. In this paper, we investigate whether we can extract these biases and transfer them into a machine recognition system. We introduce a novel method that, inspired by well-known tools in human psychophysics, estimates the biases that the human visual system might use for recognition, but in computer vision feature spaces. Our experiments are surprising, and suggest that classifiers from the human visual system can be transferred into a machine with some success. Since these classifiers seem to capture favorable biases in the human visual system, we further present an SVM formulation that constrains the orientation of the SVM hyperplane to agree with the bias from human visual system. Our results suggest that transferring this human bias into machines may help object recognition systems generalize across datasets and perform better when very little training data is available.