A Predictive Processing Account of Bottom-Up Visual Saliency Using Cross-Predicting Autoencoders
Beren Millidge, Richard C. Shillcock · 2018
We propose a novel predictive processing account of bottom-up visual saliency in which salience is simply the low-level prediction error between the sense-data and the predictions produced by the generative models in the brain. We test this with modelling in which we use cross-predicting deep autoencoders to create salience maps in an entirely unsupervised way. The resulting maps closely mimic experimentally derived human saliency maps and also spontaneously learn a centre bias, a robust viewing behaviour seen in human participants.