Research on 24-Hour Dense Crowd Counting and Object Detection System Based on Multimodal Image Optimization Feature Fusion

Table 3

Comparison of the different state-of-the-art methods on RGBT-CC dataset.

Model	MAE	MSE
UCNet [66]	33.96	56.31
HDFNet [67]	22.36	33.93
BBSNet [68]	19.56	32.48
MCNN [33]	21.89	37.44
SANet [69]	21.99	41.6
CSRNet [33]	20.4	35.26
Bayesian Loss [70]	18.7	32.67
MCNN + IADM [33]	19.77	30.34
SANet + IADM [33]	18.18	33.72
CSRNet + IADM [33]	17.94	30.91
Bayesian Loss + IADM [33]	15.61	28.18
Ours (RGBT-Net)	18.16	32.14