Table 6 Comparison of the quality of density maps on the SHT_A dataset [1] in terms of PNSR and SSIM [145].
and underline indicate the best and the worst performance, respectively. “\(\uparrow\)” indicates the higher the better of the results [1, 5, 6, 12, 96, 97, 100, 101, 103, 105, 123, 221, 250, 251, 260,261,262,263]
and underline indicate the best and the worst performance, respectively. “From: A survey of deep learning methods for density estimation and crowd counting
