Learning State Assessment in Online Education Based on Multiple Facial Features Detection

Table 1

Comparison of HOG and MMOD model on different groups.

Image size (pix)	Model	Actual faces	Detected faces	Detection time (s)
739 × 489	HOG	11	10	0.1805
	MMOD		11	39.4186
1389 × 687	HOG	10	10	0.5066
	MMOD		10	98.5127
998 × 532	HOG	11	10	0.3560
	MMOD		11	55.0837
867 × 515	HOG	12	10	0.2384
	MMOD		11	47.5020