A General Framework for Saliency Detection Methods

TL;DR

This paper introduces a comprehensive framework for saliency detection in images, outlining five key steps and comparing various models to facilitate future research in the field.

Contribution

It provides the first unified abstract framework for saliency detection methods, encompassing all major components and enabling systematic comparison.

Findings

Framework covers pre-processing to post-processing stages

Comparison of different saliency models at each level

Framework aids in developing new saliency detection methods

Abstract

Saliency detection is one of the most challenging problems in image analysis and computer vision. Many approaches propose different architectures based on the psychological and biological properties of the human visual attention system. However, there is still no abstract framework that summarizes the existing methods. In this paper, we offered a general framework for saliency models, which consists of five main steps: pre-processing, feature extraction, saliency map generation, saliency map combination, and post-processing. Also, we study different saliency models containing each level and compare their performance. This framework helps researchers to have a comprehensive view of studying new methods.