From 6d109965f3641056bb8164dc8450a7327e76e939 Mon Sep 17 00:00:00 2001 From: giuros01 Date: Mon, 7 Jan 2019 17:47:19 +0000 Subject: COMPMID-1691: Optimize CLDepthwiseConvolutionKernel (QASYMM8/NHWC) for 3x3 kernels (stride=1 and stride=2) Change-Id: I7d0d2dc350feeb40d253d17f9ffd5051a8fb42ef Reviewed-on: https://review.mlplatform.org/511 Reviewed-by: Gian Marco Iodice Tested-by: Arm Jenkins --- src/core/CL/CLKernelLibrary.cpp | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'src/core/CL/CLKernelLibrary.cpp') diff --git a/src/core/CL/CLKernelLibrary.cpp b/src/core/CL/CLKernelLibrary.cpp index 4635d11a3a..2176c59f94 100644 --- a/src/core/CL/CLKernelLibrary.cpp +++ b/src/core/CL/CLKernelLibrary.cpp @@ -225,9 +225,10 @@ const std::map CLKernelLibrary::_kernel_program_map = { "depthwise_convolution_3x3_stridex2_stridey2_bifrost_f16", "depthwise_convolution.cl" }, { "depthwise_convolution_3x3_stridex1_stridey1_bifrost_f32", "depthwise_convolution.cl" }, { "depthwise_convolution_3x3_stridex2_stridey2_bifrost_f32", "depthwise_convolution.cl" }, + { "depthwise_convolution_reshape_weights", "depthwise_convolution.cl" }, + { "depthwise_convolution_reshape_weights_generic", "depthwise_convolution.cl" }, { "depthwise_im2col", "depthwise_convolution.cl" }, { "depthwise_vector_to_tensor", "depthwise_convolution.cl" }, - { "depthwise_weights_reshape", "depthwise_convolution.cl" }, { "dequantization_layer", "dequantization_layer.cl" }, { "derivative", "derivative.cl" }, { "dilate", "dilate.cl" }, -- cgit v1.2.1