diff options
| author | Georgios Pinitas <georgios.pinitas@arm.com> | 2019-07-10 19:49:11 +0100 |
|---|---|---|
| committer | Georgios Pinitas <georgios.pinitas@arm.com> | 2019-07-12 11:07:25 +0000 |
| commit | 4c7585178385241f87288b7903d760d4b4822c6e (patch) | |
| tree | 529b6c605d29b3c1fea256558acfa3dbdc5e650e /src/core/NEON/kernels/convolution/depthwise/depthwise_dilated.cpp | |
| parent | 98b8511fd2e6fe850344ee2c69a93475bced811d (diff) | |
| download | ComputeLibrary-4c7585178385241f87288b7903d760d4b4822c6e.tar.gz | |
COMPMID-2421: Add 5x5 DepthwiseConv2d in the optimized path for NEON
Change-Id: Idecc8e7ce4404af83f0f99ed993be8f961f7b9af
Signed-off-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-on: https://review.mlplatform.org/c/1527
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/core/NEON/kernels/convolution/depthwise/depthwise_dilated.cpp')
| -rw-r--r-- | src/core/NEON/kernels/convolution/depthwise/depthwise_dilated.cpp | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/src/core/NEON/kernels/convolution/depthwise/depthwise_dilated.cpp b/src/core/NEON/kernels/convolution/depthwise/depthwise_dilated.cpp index c4f398e199..648105593c 100644 --- a/src/core/NEON/kernels/convolution/depthwise/depthwise_dilated.cpp +++ b/src/core/NEON/kernels/convolution/depthwise/depthwise_dilated.cpp @@ -30,9 +30,13 @@ template class depthwise::DilatedDepthwiseConvolution<3, 3, 3, 3, 1, 1, float, f template class depthwise::DilatedDepthwiseConvolution<3, 3, 3, 3, 2, 2, float, float, float>; template class depthwise::DilatedDepthwiseConvolution<4, 4, 3, 3, 1, 1, float, float, float>; template class depthwise::DilatedDepthwiseConvolution<4, 4, 3, 3, 2, 2, float, float, float>; +template class depthwise::DilatedDepthwiseConvolution<4, 4, 5, 5, 1, 1, float, float, float>; +template class depthwise::DilatedDepthwiseConvolution<3, 3, 5, 5, 2, 2, float, float, float>; #ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC template class depthwise::DilatedDepthwiseConvolution<3, 3, 3, 3, 1, 1, float16_t, float16_t, float16_t>; template class depthwise::DilatedDepthwiseConvolution<3, 3, 3, 3, 2, 2, float16_t, float16_t, float16_t>; +template class depthwise::DilatedDepthwiseConvolution<3, 3, 5, 5, 1, 1, float16_t, float16_t, float16_t>; +template class depthwise::DilatedDepthwiseConvolution<3, 3, 5, 5, 2, 2, float16_t, float16_t, float16_t>; #endif // __ARM_FEATURE_FP16_VECTOR_ARITHMETIC |