Table 5.

Five-fold cross validation results of different deep learning models.

Models	Index	One fold	Two fold	Three fold	Four fold	Five fold	t-test	t-test
IOCFormer	RMSE	4.69	5.18	4.79	5.10	5.34	23.25**	37.12**
	MAE	3.27	3.69	3.40	3.71	3.75	17.69**	24.40**
CSRNet	RMSE	7.49	7.82	8.18	8.13	7.92	-8.94**	1.73
	MAE	6.10	6.55	7.10	6.89	6.48	-3.93*	4.87**
SCAR	RMSE	3.36	3.45	3.25	4.12	3.79	29.79**	34.97**
	MAE	2.83	2.95	2.79	3.51	3.15	41.75**	46.46**
MCNN	RMSE	14.62	15.32	13.81	15.71	14.58	8.35**	17.86**
	MAE	13.72	14.25	13.15	14.51	13.52	12.30**	20.94**
P2PNet	RMSE	6.17	6.28	6.57	6.37	7.18	-9.97**	2.00
	MAE	4.83	4.96	5.36	5.44	5.49	-5.98**	1.63
DM-Count	RMSE	3.10	3.91	3.51	3.81	3.68	/	13.99**
	MAE	2.38	3.22	2.44	2.51	2.71	/	12.06**
DM_IOC_fpn	RMSE	2.91	3.15	3.37	3.70	3.46	-12.63**	/
	MAE	2.23	2.48	2.39	2.41	2.51	-23.80**	/