Psychophysical Machine Learning

TL;DR

This paper introduces a novel algorithm that integrates the Weber-Fechner law into machine learning loss functions, improving deep learning performance by aligning models with human perceptual principles.

Contribution

It presents a new method to incorporate psychophysical laws into loss functions, bridging human perception theories and machine learning optimization.

Findings

Enhanced deep learning performance with psychophysical loss functions

Demonstrated effectiveness across multiple neural network architectures

Provides a framework for perception-aligned machine learning models

Abstract

The Weber Fechner Law of psychophysics observes that human perception is logarithmic in the stimulus. We present an algorithm for incorporating the Weber Fechner law into loss functions for machine learning, and use the algorithm to enhance the performance of deep learning networks.