COMPMID-464 Implement Depthwise convolution 3x3 on NEON
Change-Id: Ie4e1803a52afac6b6c597c6e551729dad2347cd1
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/92607
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
diff --git a/arm_compute/core/NEON/kernels/NEDepthwiseConvolution3x3Kernel.h b/arm_compute/core/NEON/kernels/NEDepthwiseConvolution3x3Kernel.h
new file mode 100644
index 0000000..14b774c
--- /dev/null
+++ b/arm_compute/core/NEON/kernels/NEDepthwiseConvolution3x3Kernel.h
@@ -0,0 +1,69 @@
+/*
+ * 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_NEDEPTHWISECONVOLUTIONKERNEL3x3_H__
+#define __ARM_COMPUTE_NEDEPTHWISECONVOLUTIONKERNEL3x3_H__
+
+#include "arm_compute/core/NEON/INEKernel.h"
+
+namespace arm_compute
+{
+class ITensor;
+
+/** Interface for the kernel to run a 3x3 depthwise convolution on a tensor.
+ */
+class NEDepthwiseConvolution3x3Kernel : public INEKernel
+{
+public:
+ /** Default constructor */
+ NEDepthwiseConvolution3x3Kernel();
+ /** Prevent instances of this class from being copied (As this class contains pointers) */
+ NEDepthwiseConvolution3x3Kernel(const NEDepthwiseConvolution3x3Kernel &) = delete;
+ /** Prevent instances of this class from being copied (As this class contains pointers) */
+ NEDepthwiseConvolution3x3Kernel &operator=(const NEDepthwiseConvolution3x3Kernel &) = delete;
+ /** Default Move Constructor. */
+ NEDepthwiseConvolution3x3Kernel(NEDepthwiseConvolution3x3Kernel &&) = default;
+ /** Default move assignment operator. */
+ NEDepthwiseConvolution3x3Kernel &operator=(NEDepthwiseConvolution3x3Kernel &&) = 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. This is a 3D tensor 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 ITensor *input, ITensor *output, const ITensor *weights, const PadStrideInfo &conv_info);
+
+ // Inherited methods overridden:
+ void run(const Window &window, const ThreadInfo &info) override;
+ BorderSize border_size() const override;
+
+private:
+ BorderSize _border_size;
+ const ITensor *_input;
+ ITensor *_output;
+ const ITensor *_weights;
+ PadStrideInfo _conv_info;
+};
+} // namespace arm_compute
+#endif /* __ARM_COMPUTE_NEDEPTHWISECONVOLUTIONKERNEL3x3_H__ */
\ No newline at end of file