Table 1.

Summary of the VGG16 [30] architecture alongside the resulting feature vector dimension with respect to the intra-layer correlation.

Block	Layer (Name)	Layer (Type)	Kernel Size	# Filters	Feature Vector Dimension
1	conv1_1	Convolutional	$3\times 3$	64	2016
	conv1_2	Convolutional	$3\times 3$	64	2016
	Max_Pooling	Pooling	-	-	-
2	conv2_1	Convolutional	$3\times 3$	128	8128
	conv2_2	Convolutional	$3\times 3$	128	8128
	Max_Pooling	Pooling	-	-	-
3	conv3_1	Convolutional	$3\times 3$	256	32,640
	conv3_2	Convolutional	$3\times 3$	256	32,640
	conv3_3	Convolutional	$3\times 3$	256	32,640
	Max_Pooling	Pooling	-	-	-
4	conv4_1	Convolutional	$3\times 3$	512	130,816
	conv4_2	Convolutional	$3\times 3$	512	130,816
	conv4_3	Convolutional	$3\times 3$	512	130,816
	Max_Pooling	Pooling	-	-	-
5	conv5_1	Convolutional	$3\times 3$	512	130,816
	conv5_2	Convolutional	$3\times 3$	512	130,816
	conv5_3	Convolutional	$3\times 3$	512	130,816
	Max_Pooling	Pooling	-	-	-
6	fc6	Dense
7	fc7	Dense