From e52a3000d2c13bc1b66ca66b3d12b6b836982394 Mon Sep 17 00:00:00 2001 From: Gian Marco Iodice Date: Wed, 11 Apr 2018 15:59:10 +0100 Subject: COMPMID-1026 - Add support for 4x4 output tile in CLWinogradConvolutionLayer The performance achieved can be found at the following confluence page: https://confluence.arm.com/display/MLENG/GEMM-based+convolution+vs+Winograd-based+convolution+on+OpenCL Change-Id: I4b690cfdd4eb4ff0cd17b14fdd49ccaa1d1dc85c Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/127729 Tested-by: Jenkins Reviewed-by: Georgios Pinitas --- src/core/CL/CLKernelLibrary.cpp | 2 ++ 1 file changed, 2 insertions(+) (limited to 'src/core/CL/CLKernelLibrary.cpp') diff --git a/src/core/CL/CLKernelLibrary.cpp b/src/core/CL/CLKernelLibrary.cpp index 50f623fffb..59be956ad8 100644 --- a/src/core/CL/CLKernelLibrary.cpp +++ b/src/core/CL/CLKernelLibrary.cpp @@ -363,7 +363,9 @@ const std::map CLKernelLibrary::_kernel_program_map = { "winograd_input_transform_4x4_5x5_stepz1_nchw", "winograd.cl" }, { "winograd_input_transform_2x2_3x3_stepz1_nchw", "winograd.cl" }, { "winograd_input_transform_2x2_3x3_stepz2_nchw", "winograd.cl" }, + { "winograd_input_transform_4x4_3x3_stepz1_nchw", "winograd.cl" }, { "winograd_output_transform_2x2_3x3_nchw", "winograd.cl" }, + { "winograd_output_transform_4x4_3x3_nchw", "winograd.cl" }, { "winograd_output_transform_4x4_5x5_nchw", "winograd.cl" }, { "YUYV422_to_IYUV_bt709", "color_convert.cl" }, { "YUYV422_to_NV12_bt709", "color_convert.cl" }, -- cgit v1.2.1