patch for rebase master on release/1.7.0submit/tizen/20200731.060745submit/tizen/20200730.075407submit/tizen/20200730.023729accepted/tizen/unified/20200803.122223tizen_6_m1_backup

Change-Id: Id38b617d325ef7e854995a47f032bdf482a779b3

12 files changed, 33 insertions, 950 deletions

diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/NEFunctionsEx.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/NEFunctionsEx.h
index 49504fde3..3fad230f1 100644
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/NEFunctionsEx.h
+++ b/compute/ARMComputeEx/arm_compute/runtime/NEON/NEFunctionsEx.h
@@ -18,20 +18,13 @@
 
 #include <arm_compute/runtime/NEON/functions/NEActivationLayerEx.h>
 #include <arm_compute/runtime/NEON/functions/NEBinaryLogicalOperation.h>
-#include <arm_compute/runtime/NEON/functions/NECast.h>
-#include <arm_compute/runtime/NEON/functions/NEDepthToSpaceLayerEx.h>
 #include <arm_compute/runtime/NEON/functions/NEEmbeddingLookup.h>
 #include <arm_compute/runtime/NEON/functions/NEFullyConnectedReshapingLayer.h>
 #include <arm_compute/runtime/NEON/functions/NEGatherEx.h>
 #include <arm_compute/runtime/NEON/functions/NEHashtableLookup.h>
 #include <arm_compute/runtime/NEON/functions/NEInstanceNormalizationLayerEx.h>
-#include <arm_compute/runtime/NEON/functions/NEPReLU.h>
-#include <arm_compute/runtime/NEON/functions/NEReduceMeanEx.h>
 #include <arm_compute/runtime/NEON/functions/NEReduceSum.h>
-#include <arm_compute/runtime/NEON/functions/NERNNLayerEx.h>
 #include <arm_compute/runtime/NEON/functions/NEReduceOperation.h>
-#include <arm_compute/runtime/NEON/functions/NESpaceToBatchLayerEx.h>
-#include <arm_compute/runtime/NEON/functions/NESpaceToDepthLayerEx.h>
 #include <arm_compute/runtime/NEON/functions/NETransposeConvLayer.h>
 
 #endif // __ARM_COMPUTE_NEFUNCTIONSEX_H__
diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NECast.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NECast.h
deleted file mode 100644
index f0f0d8114..000000000
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NECast.h
+++ /dev/null
@@ -1,79 +0,0 @@
-/*
- * Copyright (c) 2019 Samsung Electronics Co., Ltd. All Rights Reserved
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-/*
- * Copyright (c) 2017-2019 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_NECAST_H__
-#define __ARM_COMPUTE_NECAST_H__
-
-#include "arm_compute/runtime/NEON/INESimpleFunctionNoBorder.h"
-
-#include "arm_compute/core/Types.h"
-#include "arm_compute/core/TypesEx.h"
-
-namespace arm_compute
-{
-// Forward declarations
-class ITensor;
-
-/** Basic function to run @ref NECastKernel that converts an input tensor to the other types */
-class NECast : public INESimpleFunctionNoBorder
-{
-public:
-  /** Configure the kernel.
-   *
-   * @param[in]  input  Source tensor. Data types supported: U8/S8/QASYMM8/U32/S32/F32.
-   * @param[out] output Destination tensor with the same dimensions of input. Data type supported:
-   * U8/S8/QASYMM8/U32/S32/F32.
-   * @param[in]  input_subtype  Sub data type of input.
-   */
-  void configure(const ITensor *input, ITensor *output,
-                 SubDataType input_subtype = SubDataType::NONE);
-  /** Static function to check if given info will lead to a valid configuration of @ref NECast
-   *
-   * @param[in] input  Input tensor info. Data types supported: U8/S8/QASYMM8/U32/S32/F32.
-   * @param[in] output Output tensor info. Data type supported: U8/S8/QASYMM8/U32/S32/F32.
-   * @param[in] input_subtype  Sub data type of input.
-   *
-   * @return a status
-   */
-  static Status validate(const ITensorInfo *input, const ITensorInfo *output,
-                         SubDataType input_subtype = SubDataType::NONE);
-};
-} // namespace arm_compute
-#endif /* __ARM_COMPUTE_NECAST_H__ */
diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEDepthToSpaceLayerEx.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEDepthToSpaceLayerEx.h
deleted file mode 100644
index 005d85add..000000000
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEDepthToSpaceLayerEx.h
+++ /dev/null
@@ -1,78 +0,0 @@
-/*
- * Copyright (c) 2019 Samsung Electronics Co., Ltd. All Rights Reserved
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-/*
- * Copyright (c) 2019 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_NEDEPTHTOSPACELAYEREX_H__
-#define __ARM_COMPUTE_NEDEPTHTOSPACELAYEREX_H__
-
-#include "arm_compute/runtime/IFunction.h"
-
-#include "arm_compute/core/Types.h"
-#include "arm_compute/runtime/NEON/INESimpleFunctionNoBorder.h"
-
-namespace arm_compute
-{
-class ITensor;
-
-/** Basic function to run @ref NEDepthToSpaceLayerKernelEx. */
-class NEDepthToSpaceLayerEx : public INESimpleFunctionNoBorder
-{
-public:
-  /** Set the input and output tensors.
-   *
-   * @param[in]  input       Tensor input. Supported tensor rank: 4. Data types supported:
-   * U8/S8/QASYMM8/U16/S16/F16/U32/S32/F32.
-   * @param[out] output      Tensor output. Data types supported: same as @p input
-   * @param[in]  block_shape Block shape value.
-   */
-  void configure(const ITensor *input, ITensor *output, int32_t block_shape);
-  /** Static function to check if given info will lead to a valid configuration of @ref
-   * NEDepthToSpaceLayerEx.
-   *
-   * @param[in] input       Tensor input info. Supported tensor rank: 4. Data types supported:
-   * U8/S8/QASYMM8/U16/S16/F16/U32/S32/F32.
-   * @param[in] output      Tensor output info. Data types supported: same as @p input
-   * @param[in] block_shape Block shape x value.
-   *
-   * @return a status
-   */
-  static Status validate(const ITensorInfo *input, const ITensorInfo *output, int32_t block_shape);
-};
-} // namespace arm_compute
-#endif /* __ARM_COMPUTE_NEDEPTHTOSPACELAYEREX_H__ */
diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEElementwiseUnaryLayerEx.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEElementwiseUnaryLayerEx.h
deleted file mode 100644
index 27a38e982..000000000
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEElementwiseUnaryLayerEx.h
+++ /dev/null
@@ -1,70 +0,0 @@
-/*
- * Copyright (c) 2019 Samsung Electronics Co., Ltd. All Rights Reserved
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-/*
- * Copyright (c) 2018-2019 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_NEELEMENTWISEUNARYLAYEREX_H__
-#define __ARM_COMPUTE_NEELEMENTWISEUNARYLAYEREX_H__
-
-#include "arm_compute/runtime/NEON/INESimpleFunction.h"
-
-namespace arm_compute
-{
-class ITensor;
-
-/** Basic function to perform negative on an input tensor. */
-class NENegLayer : public INESimpleFunction
-{
-public:
-  /** Initialize the function
-   *
-   * @param[in]  input  Input tensor. Data types supported: F16/F32/S32.
-   * @param[out] output Output tensor. Data types supported: same as @p input.
-   */
-  void configure(const ITensor *input, ITensor *output);
-  /** Static function to check if given info will lead to a valid configuration of @ref NERsqrtLayer
-   *
-   * @param[in] input  First tensor input info. Data types supported: F16/F32/S32.
-   * @param[in] output Output tensor info. Data types supported: Same as @p input.
-   *
-   * @return a status
-   */
-  static Status validate(const ITensorInfo *input, const ITensorInfo *output);
-};
-} // namespace arm_compute
-#endif /* __ARM_COMPUTE_NEELEMENTWISEUNARYLAYEREX_H__ */
diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEFullyConnectedHybridLayer.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEFullyConnectedHybridLayer.h
index 39c57eb70..56548a479 100644
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEFullyConnectedHybridLayer.h
+++ b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEFullyConnectedHybridLayer.h
@@ -46,7 +46,7 @@
 #include "arm_compute/core/NEON/kernels/NEMuliplyScaleFactorKernel.h"
 #include "arm_compute/core/NEON/kernels/NETransposeKernel.h"
 #include "arm_compute/runtime/MemoryGroup.h"
-#include "arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCoreEx.h"
+#include "arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCore.h"
 #include "arm_compute/runtime/NEON/INESimpleFunctionNoBorder.h"
 #include "arm_compute/runtime/Tensor.h"
 
@@ -164,7 +164,7 @@ private:
   MemoryGroup _memory_group;
   NEFullyConnectedHybridLayerReshapeWeights _reshape_weights_function;
   NEQuantizationSymmetricKernel _quant_input_kernel;
-  NEGEMMLowpMatrixMultiplyCoreEx _mm_gemmlowp;
+  NEGEMMLowpMatrixMultiplyCore _mm_gemmlowp;
   NEMultiplyScaleFactorKernel _multiply_scale_kernel;
   NEGEMMMatrixAccumulateBiasesKernel _accumulate_biases_kernel;
   Tensor _reshape_weights_output;
diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCoreEx.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCoreEx.h
deleted file mode 100644
index d844513c9..000000000
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEGEMMLowpMatrixMultiplyCoreEx.h
+++ /dev/null
@@ -1,170 +0,0 @@
-/*
- * Copyright (c) 2019 Samsung Electronics Co., Ltd. All Rights Reserved
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-/*
- * Copyright (c) 2017-2019 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_NEGEMMLOWPMATRIXMULTIPLYCOREEX_H__
-#define __ARM_COMPUTE_NEGEMMLOWPMATRIXMULTIPLYCOREEX_H__
-
-#include "arm_compute/core/NEON/INEKernel.h"
-#include "arm_compute/core/NEON/kernels/NEGEMMLowpOffsetContributionKernel.h"
-#include "arm_compute/core/NEON/kernels/NEGEMMLowpOffsetContributionOutputStageKernel.h"
-#include "arm_compute/core/NEON/kernels/NEGEMMLowpReductionKernel.h"
-#include "arm_compute/runtime/IFunction.h"
-#include "arm_compute/runtime/IMemoryManager.h"
-#include "arm_compute/runtime/MemoryGroup.h"
-// #include "arm_compute/runtime/NEON/functions/NEActivationLayer.h"
-#include "arm_compute/runtime/NEON/functions/NEGEMMAssemblyDispatch.h"
-#include "arm_compute/runtime/Tensor.h"
-
-#include <memory>
-
-namespace arm_compute
-{
-class ITensor;
-
-/** Basic function to execute GEMMLowpMatrixMultiplyCore on NEON. This function calls the following
- * NEON kernels if the DOT product instruction is not available:
- *
- *  -# @ref NEGEMMInterleave4x4Kernel
- *  -# @ref NEGEMMTranspose1xWKernel
- *  -# @ref NEGEMMLowpMatrixMultiplyKernel
- *  -# @ref NEGEMMLowpOffsetContributionKernel
- *  -# @ref NEActivationLayer
- *
- * otherwise if the DOT product instruction is available:
- *
- *  -# @ref NEGEMMLowpOffsetContributionKernel
- *
-*/
-class NEGEMMLowpMatrixMultiplyCoreEx : public IFunction
-{
-public:
-  /** Constructor */
-  NEGEMMLowpMatrixMultiplyCoreEx(std::shared_ptr<IMemoryManager> memory_manager = nullptr);
-  /** Prevent instances of this class from being copied (As this class contains pointers) */
-  NEGEMMLowpMatrixMultiplyCoreEx(const NEGEMMLowpMatrixMultiplyCoreEx &) = delete;
-  /** Default move constructor */
-  NEGEMMLowpMatrixMultiplyCoreEx(NEGEMMLowpMatrixMultiplyCoreEx &&) = default;
-  /** Prevent instances of this class from being copied (As this class contains pointers) */
-  NEGEMMLowpMatrixMultiplyCoreEx &operator=(const NEGEMMLowpMatrixMultiplyCoreEx &) = delete;
-  /** Default move assignment operator */
-  NEGEMMLowpMatrixMultiplyCoreEx &operator=(NEGEMMLowpMatrixMultiplyCoreEx &&) = default;
-  /** Initialise the kernel's inputs, output
-   *
-   * @note GEMM_LOWP:  low precision GEMM kernel
-   *  This kernel performs the following computations:
-   *
-   *  -# Convert a values from QASYMM8 to int32 and add a_offset to each of them.
-   *  -# Convert b values from QASYMM8 to int32 add b_offset to each of them.
-   *  -# Compute the matrix product of the resulting a * b in int32.
-   *
-   * @note The @p output type is S32 if @p gemm_info.type == GEMMLowpOutputStageType::NONE. It is
-   * QASYMM8/QASYMM8_SIGNED otherwise
-   *
-   * @param[in]  a         First input tensor  (Matrix A). Data type supported:
-   * QASYMM8/QASYMM8_SIGNED.
-   * @param[in]  b         Second input tensor (Matrix B). Data type supported: same as @p a
-   * @param[in]  c         Third input tensor  (Matrix C). It can be a nullptr. Data type supported:
-   * S32
-   * @param[out] output    Output tensor. Data type supported: Data type supported:
-   * S32/QASYMM8/QASYMM8_SIGNED
-   * @param[in]  gemm_info (Optional) Specifies if the matrix A and/or matrix B have been reshaped
-   * and
-   *                       if the reshape of matrix B should be executed only for the first run
-   */
-  void configure(const ITensor *a, const ITensor *b, const ITensor *c, ITensor *output,
-                 const GEMMInfo &gemm_info = GEMMInfo());
-  /** Static function to check if given info will lead to a valid configuration of @ref
-   * NEGEMMLowpMatrixMultiplyCoreEx
-   *
-   * @note The @p output type is S32 if @p gemm_info.type == GEMMLowpOutputStageType::NONE. It is
-   * QASYMM8/QASYMM8_SIGNED otherwise
-   *
-   * @param[in] a         First input tensor info  (Matrix A). Data type supported:
-   * QASYMM8/QASYMM8_SIGNED.
-   * @param[in] b         Second input tensor info (Matrix B). Data type supported: same as @p a
-   * @param[in] c         Third input tensor  info (Matrix C). It can be a nullptr. Data type
-   * supported: S32
-   * @param[in] output    Output tensor info. Data type supported: Data type supported:
-   * S32/QASYMM8/QASYMM8_SIGNED
-   * @param[in] gemm_info (Optional) Specifies if the matrix A and/or matrix B have been reshaped
-   * and
-   *                      if the reshape of matrix B should be executed only for the first run
-   *
-   * @return a status
-   */
-  static Status validate(const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *c,
-                         const ITensorInfo *output, const GEMMInfo &gemm_info = GEMMInfo());
-
-  // Inherited methods overridden
-  void run() override;
-  void prepare() override;
-
-private:
-  MemoryGroup _memory_group;
-  NEGEMMAssemblyDispatch _asm_glue;
-  std::unique_ptr<INEKernel> _mm_kernel;
-  std::unique_ptr<INEKernel> _mtx_a_reshape_kernel;
-  std::unique_ptr<INEKernel> _mtx_b_reshape_kernel;
-  NEGEMMLowpMatrixAReductionKernel _mtx_a_reduction_kernel;
-  NEGEMMLowpMatrixBReductionKernel _mtx_b_reduction_kernel;
-  NEGEMMLowpOffsetContributionKernel _offset_contribution_kernel;
-  NEGEMMLowpOffsetContributionOutputStageKernel _offset_contribution_output_stage_kernel;
-
-  Tensor _vector_sum_col;
-  Tensor _vector_sum_row;
-  Tensor _tmp_a;
-  Tensor _tmp_b;
-  Tensor _mm_result_s32;
-  Tensor _signed_a;
-  Tensor _signed_output;
-  const ITensor *_original_b;
-  int32_t _a_offset;
-  int32_t _b_offset;
-
-  bool _run_vector_matrix_multiplication;
-  bool _assembly_path;
-  bool _fused_assembly_path;
-  bool _reshape_b_only_on_first_run;
-  bool _is_prepared;
-  bool _fuse_output_stage;
-  bool _flip_signedness;
-};
-} // namespace arm_compute
-#endif /*__ARM_COMPUTE_NEGEMMLOWPMATRIXMULTIPLYCOREEX_H__ */
diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEPReLU.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEPReLU.h
deleted file mode 100644
index ca8413352..000000000
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEPReLU.h
+++ /dev/null
@@ -1,63 +0,0 @@
-/*
- * Copyright (c) 2019 Samsung Electronics Co., Ltd. All Rights Reserved
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-/*
- * Copyright (c) 2018-2019 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_NEPRELU_H__
-#define __ARM_COMPUTE_NEPRELU_H__
-
-#include "arm_compute/runtime/NEON/INESimpleFunctionNoBorder.h"
-
-namespace arm_compute
-{
-class ITensor;
-
-/** Basic function to run @ref NEPReLUKernel */
-class NEPReLU : public INESimpleFunctionNoBorder
-{
-public:
-  /** Initialise the kernel's inputs and output
-   *
-   * @param[in]  input. Data types supported: QASYMM8/F32.
-   * @param[in]  alpha. Data types supported: Same as @p input.
-   * @param[out] output Output tensor. Data types supported: Same as @p input.
-   */
-  void configure(const ITensor *input, const ITensor *alpha, ITensor *output);
-};
-} // namespace arm_compute
-#endif /* __ARM_COMPUTE_NEPRELU_H__ */
diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NERNNLayerEx.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NERNNLayerEx.h
deleted file mode 100644
index 8a7b17946..000000000
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NERNNLayerEx.h
+++ /dev/null
@@ -1,130 +0,0 @@
-/*
- * Copyright (c) 2019 Samsung Electronics Co., Ltd. All Rights Reserved
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-/*
- * Copyright (c) 2018 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_NERNNLAYER_EX_H__
-#define __ARM_COMPUTE_NERNNLAYER_EX_H__
-
-#include "arm_compute/core/NEON/kernels/NEActivationLayerKernel.h"
-#include "arm_compute/core/NEON/kernels/NEArithmeticAdditionKernel.h"
-#include "arm_compute/core/NEON/kernels/NECopyKernel.h"
-
-#include "arm_compute/core/Types.h"
-#include "arm_compute/runtime/NEON/functions/NEFullyConnectedLayer.h"
-#include "arm_compute/runtime/NEON/functions/NEGEMM.h"
-
-namespace arm_compute
-{
-// Forward declarations
-class ITensor;
-
-/** Basic function to run @ref NERNNLayerEx */
-class NERNNLayerEx : public IFunction
-{
-public:
-  /** Default constructor */
-  NERNNLayerEx(std::shared_ptr<IMemoryManager> memory_manager = nullptr);
-  /** Prevent instances of this class from being copied (As this class contains pointers) */
-  NERNNLayerEx(const NERNNLayerEx &) = delete;
-  /** Default move constructor */
-  NERNNLayerEx(NERNNLayerEx &&) = default;
-  /** Prevent instances of this class from being copied (As this class contains pointers) */
-  NERNNLayerEx &operator=(const NERNNLayerEx &) = delete;
-  /** Default move assignment operator */
-  NERNNLayerEx &operator=(NERNNLayerEx &&) = default;
-  /** Initialize the function
-   *
-   * @param[in]     input             Input is a 2-D tensor of shape [input_size, batch_size]. Data
-   * types supported: F16/F32
-   * @param[in]     weights           Weights tensor of shape [input_size, num_units] that
-   * multiplies the input. Data types supported: Same as @p input
-   * @param[in]     recurrent_weights Weights tensor of shape [num_units, num_units] that multiplies
-   * the current 'state'. Data types supported: Same as @p input
-   * @param[in]     bias              Bias vector of shape [num_units]. Data types supported: Same
-   * as @p input
-   * @param[out]    output            Output tensor of shape [num_units, batch_size]. Data types
-   * supported: Same as @p input
-   * @param[in,out] hidden_state      Output tensor of shape [num_units, batch_size]. Data types
-   * supported: Same as @p input
-   * @param[in]     info              Activation layer parameter.
-   */
-  void configure(const ITensor *input, const ITensor *weights, const ITensor *recurrent_weights,
-                 const ITensor *bias, ITensor *hidden_state, ITensor *output,
-                 ActivationLayerInfo &info);
-  /** Initialize the function
-   *
-   * @param[in] input             Input is a 2-D tensor of shape [input_size, batch_size]. Data
-   * types supported: F16/F32
-   * @param[in] weights           Weights tensor of shape [input_size, num_units] that multiplies
-   * the input. Data types supported: Same as @p input
-   * @param[in] recurrent_weights Weights tensor of shape [num_units, num_units] that multiplies the
-   * current 'state'. Data types supported: Same as @p input
-   * @param[in] bias              Bias vector of shape [num_units]. Data types supported: Same as @p
-   * input
-   * @param[in] output            Output tensor of shape [num_units, batch_size]. Data types
-   * supported: Same as @p input
-   * @param[in] hidden_state      Output tensor of shape [num_units, batch_size]. Data types
-   * supported: Same as @p input
-   * @param[in] info              Activation layer parameter.
-   *
-   * @return a status
-   */
-  static Status validate(const ITensorInfo *input, const ITensorInfo *weights,
-                         const ITensorInfo *recurrent_weights, const ITensorInfo *bias,
-                         const ITensorInfo *hidden_state, const ITensorInfo *output,
-                         const ActivationLayerInfo &info);
-
-  // Inherited methods overridden:
-  void run() override;
-  void prepare() override;
-
-private:
-  MemoryGroup _memory_group;
-  NEGEMM _gemm_state_f;
-  NEArithmeticAdditionKernel _add_kernel;
-  NEActivationLayerKernel _activation_kernel;
-  NEFullyConnectedLayer _fully_connected_kernel;
-  NECopyKernel _copy_kernel;
-  Tensor _fully_connected_out;
-  Tensor _gemm_output;
-  Tensor _add_output;
-  bool _is_prepared;
-};
-} // namespace arm_compute
-#endif /* __ARM_COMPUTE_NERNNLAYER_EX_H__ */
diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEReduceMeanEx.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEReduceMeanEx.h
deleted file mode 100644
index 03ac45798..000000000
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NEReduceMeanEx.h
+++ /dev/null
@@ -1,99 +0,0 @@
-/*
- * Copyright (c) 2019 Samsung Electronics Co., Ltd. All Rights Reserved
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-/*
- * Copyright (c) 2018 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_NEON_REDUCE_MEAN_EX_H__
-#define __ARM_COMPUTE_NEON_REDUCE_MEAN_EX_H__
-
-#include "arm_compute/runtime/IFunction.h"
-
-#include "arm_compute/core/NEON/kernels/NEFillBorderKernel.h"
-#include "arm_compute/core/Types.h"
-#include "arm_compute/runtime/MemoryGroup.h"
-#include "arm_compute/runtime/NEON/functions/NEReductionOperation.h"
-#include "arm_compute/runtime/NEON/functions/NEReshapeLayer.h"
-
-namespace arm_compute
-{
-class ITensor;
-
-/** Basic function to perform reduce operation */
-class NEReduceMeanEx : public IFunction
-{
-public:
-  /** Constructor */
-  NEReduceMeanEx(std::shared_ptr<IMemoryManager> memory_manager = nullptr);
-  /** Configure kernel
-   *
-   * @note Supported tensor rank: up to 4
-   *
-   * @param[in]  input          Source tensor. Data type supported: QASYMM8/F16/F32
-   * @param[in]  reduction_axis Reduction axis vector.
-   * @param[in]  keep_dims      If positive, retains reduced dimensions with length 1.
-   * @param[out] output         Destination tensor. Data type supported: Same as @p input
-   */
-  void configure(ITensor *input, const Coordinates &reduction_axis, bool keep_dims,
-                 ITensor *output);
-
-  /** Static function to check if given info will lead to a valid configuration of @ref
-   * NEReduceMeanEx
-   *
-   * @param[in] input          Source tensor. Data type supported: QASYMM8/F16/F32
-   * @param[in] reduction_axis Reduction axis vector.
-   * @param[in] keep_dims      If positive, retains reduced dimensions with length 1.
-   * @param[in] output         Destination tensor. Data type supported: Same as @p input
-   *
-   * @return A status
-   */
-  static Status validate(const ITensorInfo *input, const Coordinates &reduction_axis,
-                         bool keep_dims, const ITensorInfo *output);
-
-  // Inherited methods overridden:
-  void run() override;
-
-private:
-  MemoryGroup _memory_group;
-  std::unique_ptr<NEReductionOperation[]> _reduction_kernels{nullptr};
-  std::unique_ptr<Tensor[]> _reduced_outs{nullptr};
-  NEReshapeLayer _reshape;
-  unsigned int _reduction_ops;
-  bool _keep_dims;
-};
-} // namespace arm_compute
-#endif /* __ARM_COMPUTE_NEON_REDUCE_MEAN_EX_H__ */
diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NESpaceToBatchLayerEx.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NESpaceToBatchLayerEx.h
deleted file mode 100644
index 3b695fbc0..000000000
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NESpaceToBatchLayerEx.h
+++ /dev/null
@@ -1,136 +0,0 @@
-/*
- * Copyright (c) 2019 Samsung Electronics Co., Ltd. All Rights Reserved
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-/*
- * Copyright (c) 2019 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_NESPACETOBATCHLAYEREX_H__
-#define __ARM_COMPUTE_NESPACETOBATCHLAYEREX_H__
-
-#include "arm_compute/runtime/IFunction.h"
-
-#include "arm_compute/core/NEON/kernels/NEMemsetKernel.h"
-#include "arm_compute/core/NEON/kernels/NESpaceToBatchLayerKernel.h"
-#include "arm_compute/core/Types.h"
-
-namespace arm_compute
-{
-class ITensor;
-
-/** Basic function to spatial divide a tensor. This function calls the following NEON
- * kernels/functions:
- *
- *  -# @ref NEMemsetKernel
- *  -# @ref NESpaceToBatchLayerKernel
- */
-class NESpaceToBatchLayerEx : public IFunction
-{
-public:
-  /** Default constructor */
-  NESpaceToBatchLayerEx();
-  /** Prevent instances of this class from being copied (As this class contains pointers) */
-  NESpaceToBatchLayerEx(const NESpaceToBatchLayerEx &) = delete;
-  /** Prevent instances of this class from being copied (As this class contains pointers) */
-  NESpaceToBatchLayerEx &operator=(const NESpaceToBatchLayerEx &) = delete;
-  /** Allow instances of this class to be moved */
-  NESpaceToBatchLayerEx(NESpaceToBatchLayerEx &&) = default;
-  /** Allow instances of this class to be moved */
-  NESpaceToBatchLayerEx &operator=(NESpaceToBatchLayerEx &&) = default;
-  /** Default destructor */
-  virtual ~NESpaceToBatchLayerEx() = default;
-  /** Set the input and output tensors.
-   *
-   * @param[in]  input       Tensor input. Supported tensor rank: 4. Data types supported:
-   * U8/S8/QASYMM8/U16/S16/F16/U32/S32/F32.
-   * @param[in]  block_shape 1-D tensor with shape [M]. Data types supported: S32
-   * @param[in]  paddings    2-D tensor with shape [2, M]. Data types supported: S32
-   * @param[out] output      Tensor output. Data types supported: same as @p input
-   */
-  void configure(const ITensor *input, const ITensor *block_shape, const ITensor *paddings,
-                 ITensor *output);
-  /** Set the input and output tensors. (Static block shape and paddings)
-   *
-   * @param[in]  input         Tensor input. Supported tensor rank: 4. Data types supported:
-   * U8/S8/QASYMM8/U16/S16/F16/U32/S32/F32.
-   * @param[in]  block_shape_x Block shape x value.
-   * @param[in]  block_shape_y Block shape y value.
-   * @param[in]  padding_left  The left padding of the output tensor.
-   * @param[in]  padding_right The right padding of the output tensor.
-   * @param[out] output        Tensor output. Data types supported: same as @p input
-   */
-  void configure(const ITensor *input, const int block_shape_x, const int block_shape_y,
-                 const Size2D &padding_left, const Size2D &padding_right, ITensor *output);
-  /** Static function to check if given info will lead to a valid configuration of @ref
-   * NESpaceToBatchLayerEx
-   *
-   * @param[in] input       Tensor input info. Supported tensor rank: 4. Data types supported:
-   * U8/S8/QASYMM8/U16/S16/F16/U32/S32/F32.
-   * @param[in] block_shape block shape tensor info with shape [M]. Data types supported: S32
-   * @param[in] paddings    paddings tensor info with shape [2, M]. Data types supported: S32
-   * @param[in] output      Tensor output info. Data types supported: same as @p input
-   *
-   * @return a status
-   */
-  static Status validate(const ITensorInfo *input, const ITensorInfo *block_shape,
-                         const ITensorInfo *paddings, const ITensorInfo *output);
-  /** Static function to check if given info will lead to a valid configuration of @ref
-   * NESpaceToBatchLayerEx (Static block shape and paddings)
-   *
-   * @param[in] input         Tensor input info. Supported tensor rank: 4. Data types supported:
-   * U8/S8/QASYMM8/U16/S16/F16/U32/S32/F32.
-   * @param[in] block_shape_x Block shape x value.
-   * @param[in] block_shape_y Block shape y value.
-   * @param[in] padding_left  The left padding of the output tensor.
-   * @param[in] padding_right The right padding of the output tensor.
-   * @param[in] output        Tensor output info. Data types supported: same as @p input
-   *
-   * @return a status
-   */
-  static Status validate(const ITensorInfo *input, const int block_shape_x, const int block_shape_y,
-                         const Size2D &padding_left, const Size2D &padding_right,
-                         const ITensorInfo *output);
-
-  // Inherited methods overridden:
-  void run() override;
-
-private:
-  NESpaceToBatchLayerKernel _space_to_batch_kernel; /**< SpaceToBatch kernel to run */
-  NEMemsetKernel _memset_kernel;                    /**< Memset kernel to run */
-  bool _has_padding;                                /**< Flag to check if the output has padding */
-};
-} // namespace arm_compute
-#endif /* __ARM_COMPUTE_NESPACETOBATCHLAYEREX_H__ */
diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NESpaceToDepthLayerEx.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NESpaceToDepthLayerEx.h
deleted file mode 100644
index 9f32616f3..000000000
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NESpaceToDepthLayerEx.h
+++ /dev/null
@@ -1,79 +0,0 @@
-/*
- * Copyright (c) 2019 Samsung Electronics Co., Ltd. All Rights Reserved
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *      http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-/*
- * Copyright (c) 2019 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_NESPACETODEPTHLAYEREX_H__
-#define __ARM_COMPUTE_NESPACETODEPTHLAYEREX_H__
-
-#include "arm_compute/core/Types.h"
-#include "arm_compute/runtime/NEON/INESimpleFunctionNoBorder.h"
-
-namespace arm_compute
-{
-class ITensor;
-
-/** This function calls the following NEON kernels/functions:
- *
- *  -# @ref NESpaceToDepthLayerKernelEx
- */
-class NESpaceToDepthLayerEx : public INESimpleFunctionNoBorder
-{
-public:
-  /** Set the input and output tensors.
-   *
-   * @param[in]  input       Tensor input. Supported tensor rank: 4. Data types supported:
-   * U8/S8/QASYMM8/U16/S16/F16/U32/S32/F32.
-   * @param[out] output      Tensor output. Data types supported: same as @p input
-   * @param[in]  block_shape Block shape value
-   */
-  void configure(const ITensor *input, ITensor *output, int32_t block_shape);
-  /** Static function to check if given info will lead to a valid configuration of @ref
-   * NESpaceToDepthLayerEx (Static block shape and paddings)
-   *
-   * @param[in] input       Tensor input info. Supported tensor rank: 4. Data types supported:
-   * U8/S8/QASYMM8/U16/S16/F16/U32/S32/F32.
-   * @param[in] output      Tensor output info. Data types supported: same as @p input
-   * @param[in] block_shape Block shape value
-   *
-   * @return a status
-   */
-  static Status validate(const ITensorInfo *input, const ITensorInfo *output, int32_t block_shape);
-};
-} // namespace arm_compute
-#endif /* __ARM_COMPUTE_NESPACETODEPTHLAYEREX_H__ */
diff --git a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NETransposeConvLayer.h b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NETransposeConvLayer.h
index 408d150d0..24ff5dac9 100644
--- a/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NETransposeConvLayer.h
+++ b/compute/ARMComputeEx/arm_compute/runtime/NEON/functions/NETransposeConvLayer.h
@@ -15,7 +15,7 @@
  */
 
 /*
- * Copyright (c) 2017-2019 ARM Limited.
+ * Copyright (c) 2017-2020 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -37,16 +37,14 @@
  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  */
-
 #ifndef __ARM_COMPUTE_NETRANSPOSECONVLAYER_H__
 #define __ARM_COMPUTE_NETRANSPOSECONVLAYER_H__
 
-#include "arm_compute/runtime/CPP/functions/CPPUpsampleEx.h"
+#include "arm_compute/runtime/CPP/functions/CPPUpsample.h"
 #include "arm_compute/runtime/NEON/functions/NEConvolutionLayer.h"
 #include "arm_compute/runtime/NEON/functions/NEDirectConvolutionLayer.h"
-#include "arm_compute/runtime/NEON/functions/NEPermute.h"
+#include "arm_compute/runtime/NEON/functions/NEReverse.h"
 
-#include "arm_compute/core/CPP/kernels/CPPFlipWeightsKernel.h"
 #include "arm_compute/core/Types.h"
 #include "arm_compute/runtime/IFunction.h"
 #include "arm_compute/runtime/IMemoryManager.h"
@@ -59,8 +57,8 @@ namespace arm_compute
 {
 /** Function to run the deconvolution layer.
  *
- * Transpose convolution Layer is the backward pass of Convolution Layer. First we transform the
- * input depending on the stride and pad info and then perfrom a 1x1
+ * Deconvolution Layer is the backward pass of Convolution Layer. First we transform the input
+ * depending on the stride and pad info and then perfrom a 1x1
  * convolution pass. Input stride defines how many zeroes we should put between each element of the
  * input, pad is the amount of padding and finaly a is a user
  * specified value where a < stride - 1 that increases the padding top and right of the input image.
@@ -81,21 +79,22 @@ namespace arm_compute
  *      kernel_x and kernel_y are the convolution sizes in x and y.
  *      stride_x and stride_y is the input stride of the first and second dimension.
  *
- * The weights used by Transpose convolution are supposed to be the same as the ones used for
- * Convolution. Therefore, it will be necessary to use the weights in the
- * reverse order to perform an actual convolution. This is achieved by using the @ref
- * CPPFlipWeightsKernel.
+ * The weights used by Deconvolution are supposed to be the same as the ones used for Convolution.
+ * Therefore, it will be necessary to use the weights in the
+ * reverse order to perform an actual convolution. This is achieved by using @ref NEReverse.
  *
  * This function calls the following NEON kernels/functions:
  *
- * -# @ref CPPUpsample
+ * -# @ref CPPUpsampleEx
  * -# @ref NEConvolutionLayer
+ * -# @ref NEPermute
+ * -# @ref NEReverse
  *
  */
 class NETransposeConvLayer : public IFunction
 {
 public:
-  /** Default constructor */
+  /** Constructor */
   NETransposeConvLayer(std::shared_ptr<IMemoryManager> memory_manager = nullptr);
 
   /** Prevent instances of this class from being copied (As this class contains pointers) */
@@ -112,37 +111,38 @@ public:
   /** Set the input, weights, biases and output tensors.
    *
    * @param[in,out] input   Input tensor. 3 lower dimensions represent a single input, and an
-   * optional 4th dimension for batch of inputs. Data types supported: F32/F16/QASYMM8.
+ * optional 4th dimension for batch of inputs. Data types supported: F32/F16/QASYMM8/QASYMM8_SIGNED.
    * @param[in]     weights The 4d weights with dimensions [width, height, IFM, OFM]. Data type
-   * supported: Same as @p input.
+ * supported: Same as @p input.
    * @param[in]     bias    Optional, ignored if NULL. The biases have one dimension. Data type
-   * supported: Data types supported: S32 for QASYMM8 input, F32 for F32 input, F16 for F16 input.
+ * supported: Data types supported: S32 for QASYMM8 and QASYMM8_SIGNED input, F32 for F32 input, F16
+ * for F16 input.
    * @param[out]    output  Output tensor. The output has the same number of dimensions as the @p
-   * input.
+ * input.
    * @param[in]     info    Contains padding and policies to be used in the deconvolution, this is
-   * decribed in @ref PadStrideInfo.
-   * @param[in]     invalid_right  The number of zeros added to right edge of the output.
-   * @param[in]     invalid_bottom The number of zeros added to top edge of the output.
+ * decribed in @ref PadStrideInfo.
+ * @param[in]     invalid_right  The number of zeros added to right edge of the output.
+ * @param[in]     invalid_bottom The number of zeros added to bottom edge of the output.
    *
    */
   void configure(ITensor *input, const ITensor *weights, const ITensor *bias, ITensor *output,
                  const PadStrideInfo &info, unsigned int invalid_right,
                  unsigned int invalid_bottom);
   /** Static function to check if given info will lead to a valid configuration of @ref
-   * NETransposeConvLayer
+ * NETransposeConvLayer
    *
    * @param[in] input   Input tensor info. 3 lower dimensions represent a single input, and an
-   * optional 4th dimension for batch of inputs. Data types supported: F32/F16/QASYMM8.
+ * optional 4th dimension for batch of inputs. Data types supported: F32/F16/QASYMM8/QASYMM8_SIGNED.
    * @param[in] weights The 4d weights info with dimensions [width, height, IFM, OFM]. Data type
-   * supported: Same as @p input.
+ * supported: Same as @p input.
    * @param[in] bias    (Optional) The biases have one dimension. Data type supported: Data types
-   * supported: S32 for QASYMM8 input, F32 for F32 input, F16 for F16 input.
+ * supported: S32 for QASYMM8 and QASYMM8_SIGNED input, F32 for F32 input, F16 for F16 input.
    * @param[in] output  Output tensor info. The output has the same number of dimensions as the @p
-   * input.
+ * input.
    * @param[in] info    Contains padding and policies to be used in the deconvolution, this is
-   * decribed in @ref PadStrideInfo.
-   * @param[in] innvalid_right  The number of zeros added to right edge of the output.
-   * @param[in] invalid_bottom  The number of zeros added to top edge of the output.
+ * decribed in @ref PadStrideInfo.
+ * @param[in] innvalid_right  The number of zeros added to right edge of the output.
+ * @param[in] invalid_bottom  The number of zeros added to bottom edge of the output.
    *
    * @return a status
    */
@@ -158,17 +158,11 @@ public:
 private:
   MemoryGroup _memory_group;
   NEConvolutionLayer _conv_f;
-  CPPUpsampleEx _upsample_f;
-  CPPFlipWeightsKernel _flip_weights;
-  NEPermute _permute_input;
-  NEPermute _permute_weights;
-  NEPermute _permute_output;
+  CPPUpsample _upsample_f;
+  NEReverse _flip_weights;
   Tensor _scaled_output;
   Tensor _weights_flipped;
-  Tensor _permuted_input;
-  Tensor _permuted_weights;
-  Tensor _permuted_output;
-  bool _is_nchw;
+  Tensor _flip_axis;
   const ITensor *_original_weights;
   ITensor *_input;
   PadStrideInfo _info;