>The rest of the paper is organized as follows. §2 explains the proposed taxonomies, each accompanied with one or two most representative models. §3 examines the most notable SOD datasets, whereas §4 describes several widely used SOD metrics. §5 benchmarks several deep SOD models and provides in-depth analyses. §6 provides further discussions and presents open issues and future research directions of the field. Finally, §7 concludes the paper