In this study, we employed stringent benchmark datasets proposed by Chen et al., which can be downloaded from (here). The details of the benchmark datasets can be referred to the related papers.