Ultra-low power and high-performance AI processor GAP8 › Forums › GAP8 developers’ group › SDK (Software Development Kit) › 1xN convolutions
- saseptimParticipantMay 2, 2019 at 9:43 amPost count: 2
I did not notice any support for non-square convolutions. In audio processing, one could use 1×3, 1×5, etc – 1D convolutions.
Do you have plans to support these convolutions? I guess one solution would be to use a 3×3 filter where only the first 3 are non-zero but it seems quite wasteful.
Also, I am looking at the basic convolution implementation and trying to understand the line:
gap8_clip(gap8_roundnorm_reg(Acc, Norm), 7)
I understand that you are shifting Acc right by Norm and then clipping to int8 (byte). Where does “Norm” come from? I see that in the mnist example, you use 14.
static void __attribute__ ((noinline)) KerConvNxNStrideS_Body_fps(
signed char *__restrict__ In,
signed char *__restrict__ Out,
signed char *__restrict__ Filter,
unsigned short int PadL = Pad, PadT = Pad;
signed char *PtO = Out+Wo*Ho_F+Wo_F;
signed char *PtC = Filter;
for ( int h=Ho_F; h
- You must be logged in to reply to this topic.