diff options
author | Giorgio Arena <giorgio.arena@arm.com> | 2017-08-23 16:36:24 +0100 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:35:24 +0000 |
commit | 9fe414430c3c989b1cdc79d41e031495aed2cb7c (patch) | |
tree | 5f893eeb81fe75824918273f49171c85f108fe4e /arm_compute/core/CL/kernels/CLDepthwiseConvolution3x3Kernel.h | |
parent | 580f4c7f6c2cbe320c606eb6ba73b3f4a20791ef (diff) | |
download | ComputeLibrary-9fe414430c3c989b1cdc79d41e031495aed2cb7c.tar.gz |
COMPMID-452 CL Generic Depthwise Convolution implementation.
Change-Id: I115e48fe6ce5e281f3791aa5d80fdc754cdd2b5e
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/85082
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Diffstat (limited to 'arm_compute/core/CL/kernels/CLDepthwiseConvolution3x3Kernel.h')
-rw-r--r-- | arm_compute/core/CL/kernels/CLDepthwiseConvolution3x3Kernel.h | 72 |
1 files changed, 72 insertions, 0 deletions
diff --git a/arm_compute/core/CL/kernels/CLDepthwiseConvolution3x3Kernel.h b/arm_compute/core/CL/kernels/CLDepthwiseConvolution3x3Kernel.h new file mode 100644 index 0000000000..4e69f551b8 --- /dev/null +++ b/arm_compute/core/CL/kernels/CLDepthwiseConvolution3x3Kernel.h @@ -0,0 +1,72 @@ +/* + * Copyright (c) 2017 ARM Limited. + * + * SPDX-License-Identifier: MIT + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in all + * copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ +#ifndef __ARM_COMPUTE_CLDEPTHWISECONVOLUTIONKERNEL3x3_H__ +#define __ARM_COMPUTE_CLDEPTHWISECONVOLUTIONKERNEL3x3_H__ + +#include "arm_compute/core/CL/ICLKernel.h" + +namespace arm_compute +{ +class ICLTensor; + +/** Interface for the kernel to run a 3x3 depthwise convolution on a tensor. + */ +class CLDepthwiseConvolution3x3Kernel : public ICLKernel +{ +public: + /** Default constructor */ + CLDepthwiseConvolution3x3Kernel(); + /** Prevent instances of this class from being copied (As this class contains pointers) */ + CLDepthwiseConvolution3x3Kernel(const CLDepthwiseConvolution3x3Kernel &) = delete; + /** Prevent instances of this class from being copied (As this class contains pointers) */ + CLDepthwiseConvolution3x3Kernel &operator=(const CLDepthwiseConvolution3x3Kernel &) = delete; + /** Default Move Constructor. */ + CLDepthwiseConvolution3x3Kernel(CLDepthwiseConvolution3x3Kernel &&) = default; + /** Default move assignment operator. */ + CLDepthwiseConvolution3x3Kernel &operator=(CLDepthwiseConvolution3x3Kernel &&) = default; + /** Initialize the function's source, destination, conv and border_size. + * + * @param[in] input Source tensor. DataType supported: F32. + * @param[out] output Destination tensor. Data type supported: Same as @p input. + * @param[in] weights Weights tensor. These are 3D tensors with dimensions [3, 3, IFM]. Data type supported: Same as @p input. + * @param[in] conv_info Padding and stride information to use for the convolution. + */ + void configure(const ICLTensor *input, ICLTensor *output, const ICLTensor *weights, const PadStrideInfo &conv_info); + + // Inherited methods overridden: + void run(const Window &window, cl::CommandQueue &queue) override; + BorderSize border_size() const override; + +private: + BorderSize _border_size; + const ICLTensor *_input; + ICLTensor *_output; + const ICLTensor *_weights; + unsigned int _conv_stride_x; + unsigned int _conv_stride_y; + unsigned int _conv_pad_x; + unsigned int _conv_pad_y; +}; +} // namespace arm_compute +#endif /*__ARM_COMPUTE_CLDEPTHWISECONVOLUTIONKERNEL3x3_H__ */ |