Table 1.

Number of learnable parameters for all models and input channels.

	B	B-F	B-K	B-D	BT	BL	BLT	GLM
Kernel size	$3\times 3$	$3\times 3$	$6\times 6$	$3\times 3$	$3\times 3$	$3\times 3$	$3\times 3$	–
Hidden units	32	64	32	32	32	32	32	–
Layers	2	2	2	4	2	2	2	1
Channels	Number of learnable parameters (OS-MNIST, 10 classes)
1	$9,898$	$38,218$	$38,410$	$28,394$	$19,146$	$28,394$	$37,642$	$10,250$
2	$10,186$	$38,794$	$39,562$	$28,682$	$19,434$	$28,682$	$37,930$	$20,490$
Channels	Number of learnable parameters (OS-YCB, 79 classes)
3	$12,751$	$43,855$	$40,714$	$31,247$	$21,999$	$31,247$	$40,495$	$242,767$
6	$13,615$	$45,583$	$44,170$	$32,111$	$22,863$	$32,111$	$41,359$	$485,455$