diff options
| author | Gian Marco Iodice <gianmarco.iodice@arm.com> | 2017-08-29 16:05:25 +0100 |
|---|---|---|
| committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:35:24 +0000 |
| commit | 06b184ac568dc974986bae680957c4477f8ef6ca (patch) | |
| tree | fa97d020f81f9a17edb6b50394f2bdf46f810ce9 /arm_compute/runtime/NEON/functions/NEDequantizationLayer.h | |
| parent | 351c20a361521101307d365a4f91ad883fa272ea (diff) | |
| download | ComputeLibrary-06b184ac568dc974986bae680957c4477f8ef6ca.tar.gz | |
COMPMID-439 - Refactored NEQuantizationLayer and NEQuantizationLayer in order to support 3D input tensors
Change-Id: I03eac2108a30bed56d40dfd52e75577a35d492e0
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/85783
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Diffstat (limited to 'arm_compute/runtime/NEON/functions/NEDequantizationLayer.h')
| -rw-r--r-- | arm_compute/runtime/NEON/functions/NEDequantizationLayer.h | 13 |
1 files changed, 7 insertions, 6 deletions
diff --git a/arm_compute/runtime/NEON/functions/NEDequantizationLayer.h b/arm_compute/runtime/NEON/functions/NEDequantizationLayer.h index 7cd8360713..898586190e 100644 --- a/arm_compute/runtime/NEON/functions/NEDequantizationLayer.h +++ b/arm_compute/runtime/NEON/functions/NEDequantizationLayer.h @@ -27,7 +27,6 @@ #include "arm_compute/runtime/IFunction.h" #include "arm_compute/core/NEON/kernels/NEDequantizationLayerKernel.h" -#include "arm_compute/runtime/Tensor.h" #include "arm_compute/core/Types.h" @@ -37,6 +36,8 @@ class ITensor; /** Basic function to simulate a dequantization layer. This function calls the following NEON kernels: * + * @note The implementation supports only 3D input tensors + * * -# @ref NEDequantizationLayerKernel * */ @@ -47,12 +48,12 @@ public: NEDequantizationLayer(); /** Configure the kernel. * - * @param[in] input Source tensor. Data types supported: U8. - * @param[out] output Destination tensor. Data types supported: F32. - * @param[in] min Minimum value of the input tensor. - * @param[in] max Maximum value of the input tensor. + * @param[in] input Source tensor with at least 3 dimensions. The dimensions over the third will be interpreted as batches. Data types supported: U8. + * @param[out] output Destination tensor with the same dimensions of input. Data type supported: F32. + * @param[in] min_max Pointer to the tensor with shape [2, batches] which stores the minimum and maximum value for each 3D input tensor. + * The dimensions over the second must match the batched dimensions of the input tensor. Data type supported: F32 */ - void configure(const ITensor *input, ITensor *output, const float *min, const float *max); + void configure(const ITensor *input, ITensor *output, const ITensor *min_max); // Inherited methods overridden: void run() override; |