diff options
Diffstat (limited to 'arm_compute/runtime/NEON/functions/NEFFTConvolutionLayer.h')
-rw-r--r-- | arm_compute/runtime/NEON/functions/NEFFTConvolutionLayer.h | 35 |
1 files changed, 26 insertions, 9 deletions
diff --git a/arm_compute/runtime/NEON/functions/NEFFTConvolutionLayer.h b/arm_compute/runtime/NEON/functions/NEFFTConvolutionLayer.h index 8967363e7f..84bfe6b02f 100644 --- a/arm_compute/runtime/NEON/functions/NEFFTConvolutionLayer.h +++ b/arm_compute/runtime/NEON/functions/NEFFTConvolutionLayer.h @@ -24,9 +24,8 @@ #ifndef ARM_COMPUTE_NEFFTCONVOLUTIONLAYER_H #define ARM_COMPUTE_NEFFTCONVOLUTIONLAYER_H -#include "arm_compute/runtime/IFunction.h" - #include "arm_compute/core/Types.h" +#include "arm_compute/runtime/IFunction.h" #include "arm_compute/runtime/NEON/functions/NEActivationLayer.h" #include "arm_compute/runtime/NEON/functions/NEArithmeticAddition.h" #include "arm_compute/runtime/NEON/functions/NEFFT2D.h" @@ -43,7 +42,7 @@ namespace arm_compute // Forward declarations class ITensor; -/** Basic function to execute FFT-based convolution on Neon. This function calls the following Neon functions/kernels: +/** Basic function to execute FFT-based convolution on CPU. This function calls the following functions/kernels: * * -# @ref NEPermute Permute input if NHWC(only NCHW is supported). * -# @ref NEPadLayer Pad input. @@ -73,6 +72,14 @@ public: ~NEFFTConvolutionLayer(); /** Set the input and output tensors. * + * Valid data layouts: + * - All + * + * Valid data type configurations: + * |src |dst | + * |:------|:------| + * |F32 |F32 | + * * @note: This function only works with any square kernel size and unit strides for both NCHW and NHWC data layout * * @param[in] input Source tensor. 3 lower dimensions represent a single input [width, height, IFM], @@ -84,10 +91,15 @@ public: * Data types supported: Same as @p input. * @param[in] conv_info Contains padding and stride information described in @ref PadStrideInfo. * @param[in] act_info (Optional) Activation layer information in case of a fused activation. - * @param[in] enable_fast_math (Optional) Enable fast math computation. Unused for Neon backend. + * @param[in] enable_fast_math (Optional) Enable fast math computation. Unused for CPU backend. */ - void configure(ITensor *input, const ITensor *weights, const ITensor *biases, ITensor *output, const PadStrideInfo &conv_info, - const ActivationLayerInfo &act_info = ActivationLayerInfo(), bool enable_fast_math = false); + void configure(ITensor *input, + const ITensor *weights, + const ITensor *biases, + ITensor *output, + const PadStrideInfo &conv_info, + const ActivationLayerInfo &act_info = ActivationLayerInfo(), + bool enable_fast_math = false); /** Static function to check if given info will lead to a valid configuration of @ref NEFFTConvolutionLayer * * @note: This function only works with any square kernel size and unit strides for both NCHW and NHWC data layout @@ -101,12 +113,17 @@ public: * Data types supported: Same as @p input. * @param[in] conv_info Contains padding and stride information described in @ref PadStrideInfo. * @param[in] act_info (Optional) Activation layer information in case of a fused activation. - * @param[in] enable_fast_math (Optional) Enable fast math computation. Unused for Neon backend. + * @param[in] enable_fast_math (Optional) Enable fast math computation. Unused for CPU backend. * * @return a status */ - static Status validate(const ITensorInfo *input, const ITensorInfo *weights, const ITensorInfo *biases, const ITensorInfo *output, const PadStrideInfo &conv_info, - const ActivationLayerInfo &act_info = ActivationLayerInfo(), bool enable_fast_math = false); + static Status validate(const ITensorInfo *input, + const ITensorInfo *weights, + const ITensorInfo *biases, + const ITensorInfo *output, + const PadStrideInfo &conv_info, + const ActivationLayerInfo &act_info = ActivationLayerInfo(), + bool enable_fast_math = false); // Inherited methods overridden: void run() override; |