This paper presents a comprehensive benchmark for salient object detection, comparing 41 state-of-the-art models across seven challenging datasets. The evaluation covers both qualitative and quantitative aspects, assessing models in terms of accuracy and running time. The results show significant progress in recent years, with top-performing models outperforming those from three years ago. The study highlights that models specifically designed for salient object detection generally outperform models in related areas, emphasizing the distinct challenges and solutions in this field. The analysis also explores the impact of center bias and scene complexity on model performance, suggesting ways to construct more challenging datasets and improve models. Additionally, the paper proposes solutions for addressing open problems such as evaluation scores and dataset bias, providing insights into future research directions in salient object detection.This paper presents a comprehensive benchmark for salient object detection, comparing 41 state-of-the-art models across seven challenging datasets. The evaluation covers both qualitative and quantitative aspects, assessing models in terms of accuracy and running time. The results show significant progress in recent years, with top-performing models outperforming those from three years ago. The study highlights that models specifically designed for salient object detection generally outperform models in related areas, emphasizing the distinct challenges and solutions in this field. The analysis also explores the impact of center bias and scene complexity on model performance, suggesting ways to construct more challenging datasets and improve models. Additionally, the paper proposes solutions for addressing open problems such as evaluation scores and dataset bias, providing insights into future research directions in salient object detection.