. 2022 Nov 24;132:109851. doi: 10.1016/j.asoc.2022.109851

Table 4.

Experimental configurations.

Conf.	Network	Batch size	# of Params	Learning method	Size (MB)
C $_{1}$	EfficientNet-B0	110	7,919,391	ImageNet	53.1
C $_{2}$	EfficientNet-B0	110	7,919,391	AdvProp	53.1
C $_{3}$	EfficientNet-B0	110	7,919,391	NS	53.1
C $_{4}$	EfficientNet-B3	64	14,352,075	ImageNet	106.9
C $_{5}$	EfficientNet-B3	64	14,352,075	AdvProp	106.9
C $_{6}$	EfficientNet-B3	64	14,352,075	NS	106.9

C $_{7}$	MixNet-Small	110	6,253,449	ImageNet	41.8
C $_{8}$	MixNet-Medium	90	7,133,225	ImageNet	48.9
C $_{9}$	MixNet-Large	60	9,448,095	ImageNet	67.5
C₁₀	MixNet-XL	60	14,015,611	ImageNet	104.2