Conv2D

2D convolution on H, W dimensions of BCHW/BHWC using kernels MCKK. Each c-th KxK kernel is applied on C dimension. This is done over M iterations to yield MxHxW per instance. This is done over B iterations to yield B batches. More...

Collaboration diagram for Conv2D:

Modules
	Conv2DKernels

Classes
class	ConvReluGraph< CONV, INP_H, INP_W, INP_W_PAD, OUT_W, OUT_W_PAD, STEP_H, STEP_W, B, C, M, KH, KW, GROUP, IS_RELU, H0, H1, W0, W1 >
	Single instance graph that stores weights and biases. More...
class	ConvReluStreamGraph< CONV, INP_H, INP_W, INP_W_PAD, OUT_W, OUT_W_PAD, STEP_H, STEP_W, B, C, M, KH, KW, GROUP, IS_RELU, H0, H1, W0, W1 >
	Single instance graph that streams weights and biases, significantly slower. More...
class	ConvReluChunkMGraph< CONV, CONCAT, IS_BCHW, MCHUNK, INP_H, INP_W, INP_W_PAD, OUT_W, OUT_W_PAD, STEP_H, STEP_W, B, C, M, KH, KW, GROUP, IS_RELU, H0, H1, W0, W1 >
	Multiinstance graph that stores weights and biases, chunks MCKK weights by M dimension, maximum 8 chunks If IS_BCHW=0 (using BHWC kernel): MCHUNK%8=0 and M%4=0. If IS_BCHW=1 (using BCHW kernel): MCHUNK*OUT_W_PAD*OUT_W_PAD%8=0 and M*OUT_W_PAD*OUT_W_PAD%4=0. More...
class	ConvReluChunkHGraph< SPLIT, CONV, CONCAT, HCHUNK, INP_H, INP_W, INP_W_PAD, OUT_W, OUT_W_PAD, STEP_H, STEP_W, B, C, M, KH, KW, GROUP, IS_RELU, H0, H1, W0, W1 >
	Multiinstance graph that stores weights and biases, chunks BCHW by H dimension, maximum 8 chunks. More...
class	ConvReluChunkHStreamGraph< SPLIT, CONV, CONCAT, HCHUNK, INP_H, INP_W, INP_W_PAD, OUT_W, OUT_W_PAD, STEP_H, STEP_W, B, C, M, KH, KW, GROUP, IS_RELU, H0, H1, W0, W1 >
	Multiinstance graph that stores biases, chunks BCHW by H dimension. More...
class	ConvReluChunkHPktStreamGraph< SPLIT, CONV, CONCAT, HCHUNK, INP_H, INP_W, INP_W_PAD, OUT_W, OUT_W_PAD, STEP_H, STEP_W, B, C, M, KH, KW, GROUP, IS_RELU, H0, H1, W0, W1 >
	Multiinstance graph that stores biases, chunks BCHW by H dimension. More...

Detailed Description

• std::conditional for kernel/graph typedef results in error in graph hierarchy algorithm

Template Parameters

CONV	Conv2D Kernel
CONCAT	Concat Kernel (if multiinstance)
IS_BCHW	if BCHW or BHWC, affects concatenation (if multiinstance)
MCHUNK	M chunk size (if multiinstance)
HCHUNK	H chunk size (if multiinstance)
INP_H	input height
INP_W	input width
INP_W_PAD	input width
OUT_W	output width
OUT_W_PAD	output width padded to vector boundary
STEP_H	stride H
STEP_W	stride W
B	batch size
C	input channels
M	output channels
KH	kernel height
KW	kernel width
GROUP	split input into groups, CGROUP==0
H0	Pixels added before height (default 0)
H1	Pixels added after height (default 0)
W0	Pixels added before width (default 0)
W1	Pixels added after width (default 0)