Saliency for free: Saliency prediction as a side-effect of object recognition

TL;DR

This paper proposes a method to generate saliency maps as a byproduct of training an object recognition neural network with a saliency branch, eliminating the need for ground-truth saliency data.

Contribution

It introduces a neural network architecture that produces saliency maps without requiring explicit saliency annotations during training.

Findings

Achieves competitive saliency prediction results on real datasets.

Works effectively on synthetic saliency datasets.

Eliminates the need for eye-tracking ground truth.

Abstract

Saliency is the perceptual capacity of our visual system to focus our attention (i.e. gaze) on relevant objects. Neural networks for saliency estimation require ground truth saliency maps for training which are usually achieved via eyetracking experiments. In the current paper, we demonstrate that saliency maps can be generated as a side-effect of training an object recognition deep neural network that is endowed with a saliency branch. Such a network does not require any ground-truth saliency maps for training.Extensive experiments carried out on both real and synthetic saliency datasets demonstrate that our approach is able to generate accurate saliency maps, achieving competitive results on both synthetic and real datasets when compared to methods that do require ground truth data.