1 /* 2 * Copyright (c) 2017-2021 Arm Limited. 3 * 4 * SPDX-License-Identifier: MIT 5 * 6 * Permission is hereby granted, free of charge, to any person obtaining a copy 7 * of this software and associated documentation files (the "Software"), to 8 * deal in the Software without restriction, including without limitation the 9 * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or 10 * sell copies of the Software, and to permit persons to whom the Software is 11 * furnished to do so, subject to the following conditions: 12 * 13 * The above copyright notice and this permission notice shall be included in all 14 * copies or substantial portions of the Software. 15 * 16 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 17 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 18 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 19 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 20 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 21 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 22 * SOFTWARE. 23 */ 24 #ifndef ARM_COMPUTE_CL_IM2COL_KERNEL_H 25 #define ARM_COMPUTE_CL_IM2COL_KERNEL_H 26 27 #include "arm_compute/core/KernelDescriptors.h" 28 #include "arm_compute/core/Size2D.h" 29 #include "src/core/common/Macros.h" 30 #include "src/gpu/cl/ClCompileContext.h" 31 #include "src/gpu/cl/IClKernel.h" 32 33 namespace arm_compute 34 { 35 namespace opencl 36 { 37 namespace kernels 38 { 39 /** Interface for the im2col reshape kernel. 40 * 41 * Rearranges image blocks into columns. It is used to strip out each convolution block to a single column. 42 * It is used to transform a convolution to a plain matrix multiplication. 43 * 44 * For example taking into account the image below and assuming 3x3 image blocks with stride of 1 we have: 45 * @f[ 46 * \left( \begin{array}{cccc} 47 * a00 & a01 & a02 & a03 \\ 48 * a10 & a11 & a12 & a13 \\ 49 * a20 & a21 & a22 & a23 \\ 50 * a30 & a31 & a32 & a33 \\ 51 * \end{array} \right) 52 * = 53 * \left( \begin{array}{ccccccccc} 54 * a00 & a01 & a02 & a10 & a11 & a12 & a20 & a21 & a22 \\ 55 * a01 & a02 & a03 & a11 & a12 & a13 & a21 & a22 & a23 \\ 56 * a10 & a11 & a12 & a20 & a21 & a22 & a30 & a31 & a32 \\ 57 * a11 & a12 & a13 & a21 & a22 & a23 & a31 & a32 & a33 \\ 58 * \end{array} \right) 59 * @f] 60 */ 61 class ClIm2ColKernel : public IClKernel 62 { 63 public: 64 /** Default constructor */ 65 ClIm2ColKernel(); 66 ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(ClIm2ColKernel); 67 /** Set the input and output of the kernel. 68 * 69 * @param[in] compile_context The compile context to be used. 70 * @param[in] src The input tensor info to convert. 3 lower dimensions represent a single input [width, height, IFM], 71 * while every optional dimension from 4 and above represent a batch of inputs. Data types supported: QASYMM8/QASYMM8_SIGNED/F16/F32 72 * @param[out] dst The output tensor info. First 2 lower dimensions represent a transform of each 3D input, 73 * while every dimension above represents a batch. Data types supported: Same as @p input 74 * @param[in] kernel_dims The kernel dimensions (width and height). 75 * @param[in] conv_info Contains padding and stride information described in @ref PadStrideInfo. 76 * @param[in] has_bias In case biases are provided expands the matrix with 1. 77 * @param[in] dilation (Optional) Dilation, in elements, across x and y. Defaults to (1, 1). 78 * @param[in] num_groups (Optional) Number of groups when performing a grouped convolution. num_groups != 1 is only supported for NCHW data layout 79 */ 80 void configure(const ClCompileContext &compile_context, ITensorInfo *src, ITensorInfo *dst, const Size2D &kernel_dims, const PadStrideInfo &conv_info, bool has_bias, 81 const Size2D &dilation = Size2D(1U, 1U), 82 unsigned int num_groups = 1); 83 /** Static function to check if given info will lead to a valid configuration 84 * 85 * Similar to ClIm2ColKernel::configure() 86 * 87 * @return a status 88 */ 89 static Status validate(const ITensorInfo *input, const ITensorInfo *output, const Size2D &kernel_dims, const PadStrideInfo &conv_info, bool has_bias, const Size2D &dilation = Size2D(1U, 1U), 90 unsigned int num_groups = 1); 91 92 // Inherited methods overridden: 93 void run_op(ITensorPack &tensors, const Window &window, cl::CommandQueue &queue) override; 94 95 public: 96 DataLayout _data_layout; 97 std::pair<unsigned int, unsigned int> _convolved_dims; 98 unsigned int _num_elems_processed_per_iteration; 99 Size2D _kernel_dims; 100 PadStrideInfo _conv_info; 101 unsigned int _num_groups; 102 }; 103 } // namespace kernels 104 } // namespace opencl 105 } // namespace arm_compute 106 #endif /* ARM_COMPUTE_CL_IM2COL_KERNEL_H */ 107