diff options
author | giuros01 <giuseppe.rossini@arm.com> | 2019-01-07 17:47:19 +0000 |
---|---|---|
committer | Giuseppe Rossini <giuseppe.rossini@arm.com> | 2019-01-30 16:22:47 +0000 |
commit | 6d109965f3641056bb8164dc8450a7327e76e939 (patch) | |
tree | 45e40a75e7f2d80e403a33087284f08b2b2a0b6b /arm_compute/core/CL/kernels/CLDepthwiseConvolutionLayer3x3NHWCKernel.h | |
parent | edc21e44313edea693700a6bdfa353edcfbe25be (diff) | |
download | ComputeLibrary-6d109965f3641056bb8164dc8450a7327e76e939.tar.gz |
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 <gianmarco.iodice@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'arm_compute/core/CL/kernels/CLDepthwiseConvolutionLayer3x3NHWCKernel.h')
-rw-r--r-- | arm_compute/core/CL/kernels/CLDepthwiseConvolutionLayer3x3NHWCKernel.h | 11 |
1 files changed, 6 insertions, 5 deletions
diff --git a/arm_compute/core/CL/kernels/CLDepthwiseConvolutionLayer3x3NHWCKernel.h b/arm_compute/core/CL/kernels/CLDepthwiseConvolutionLayer3x3NHWCKernel.h index 85fbaaee37..2fc9780a2f 100644 --- a/arm_compute/core/CL/kernels/CLDepthwiseConvolutionLayer3x3NHWCKernel.h +++ b/arm_compute/core/CL/kernels/CLDepthwiseConvolutionLayer3x3NHWCKernel.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018 ARM Limited. + * Copyright (c) 2018-2019 ARM Limited. * * SPDX-License-Identifier: MIT * @@ -53,11 +53,11 @@ public: ActivationLayerInfo act_info) override; /** Static function to check if given info will lead to a valid configuration of @ref CLDepthwiseConvolutionLayer3x3NHWCKernel * - * @param[in] input Source tensor. DataType supported: QASYMM8. - * @param[in] weights Weights tensor. A 3D tensor with dimensions [IFM, 3, 3]. Data type supported: Same as @p input. - * @param[in] biases (Optional) Biases tensor. A 1D tensor with dimensions [IFM]. Must be nullptr if not needed. + * @param[in] input Source tensor info. DataType supported: QASYMM8. + * @param[in] weights Weights tensor info. A 3D tensor with dimensions [IFM, 3, 3]. Data type supported: Same as @p input. + * @param[in] biases (Optional) Biases tensor info. A 1D tensor with dimensions [IFM]. Must be nullptr if not needed. * Data type supported: Same as @p input. - * @param[in] output Destination tensor. Data type supported: Same as @p input. + * @param[in] output Destination tensor info. Data type supported: Same as @p input. * @param[in] conv_info Padding and stride information to use for the convolution. * @param[in] depth_multiplier (Optional) Multiplier to apply to the input's depth in order to retrieve the output's depth. Defaults to 1. * @param[in] act_info (Optional) Activation layer information in case of a fused activation. Only RELU, BOUNDED_RELU and LU_BOUNDED_RELU are supported. @@ -67,6 +67,7 @@ public: static Status validate(const ITensorInfo *input, const ITensorInfo *weights, const ITensorInfo *biases, const ITensorInfo *output, const PadStrideInfo &conv_info, unsigned int depth_multiplier, ActivationLayerInfo act_info = ActivationLayerInfo()); + // Inherited methods overridden: void run(const Window &window, cl::CommandQueue &queue) override; BorderSize border_size() const override; |