diff --git a/tests/CL/Helper.h b/tests/CL/Helper.h
deleted file mode 100644
index a6063e9..0000000
--- a/tests/CL/Helper.h
+++ /dev/null
@@ -1,76 +0,0 @@
-/*
- * 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_TEST_CL_HELPER_H__
-#define __ARM_COMPUTE_TEST_CL_HELPER_H__
-
-#include "Globals.h"
-#include "TensorLibrary.h"
-
-#include "arm_compute/runtime/CL/CLTensor.h"
-
-namespace arm_compute
-{
-namespace test
-{
-namespace cl
-{
-/** Helper to create an empty tensor.
- *
- * @param[in] shape                Desired shape.
- * @param[in] data_type            Desired data type.
- * @param[in] num_channels         (Optional) It indicates the number of channels for each tensor element
- * @param[in] fixed_point_position (Optional) Fixed point position that expresses the number of bits for the fractional part of the number when the tensor's data type is QS8 or QS16.
- *
- * @return Empty @ref CLTensor with the specified shape and data type.
- */
-inline CLTensor create_tensor(const TensorShape &shape, DataType data_type, int num_channels = 1, int fixed_point_position = 0)
-{
-    CLTensor tensor;
-    tensor.allocator()->init(TensorInfo(shape, num_channels, data_type, fixed_point_position));
-
-    return tensor;
-}
-
-/** Helper to create an empty tensor.
- *
- * @param[in] name      File name from which to get the dimensions.
- * @param[in] data_type Desired data type.
- *
- * @return Empty @ref CLTensor with the specified shape and data type.
- */
-inline CLTensor create_tensor(const std::string &name, DataType data_type)
-{
-    constexpr unsigned int num_channels = 1;
-
-    const RawTensor &raw = library->get(name);
-
-    CLTensor tensor;
-    tensor.allocator()->init(TensorInfo(raw.shape(), num_channels, data_type));
-
-    return tensor;
-}
-} // namespace cl
-} // namespace test
-} // namespace arm_compute
-#endif
diff --git a/tests/NEON/Helper.h b/tests/NEON/Helper.h
index e64ac27..e776154 100644
--- a/tests/NEON/Helper.h
+++ b/tests/NEON/Helper.h
@@ -24,11 +24,9 @@
 #ifndef __ARM_COMPUTE_TEST_NEON_HELPER_H__
 #define __ARM_COMPUTE_TEST_NEON_HELPER_H__
 
-#include "Globals.h"
-#include "TensorLibrary.h"
-
 #include "arm_compute/runtime/Array.h"
-#include "arm_compute/runtime/Tensor.h"
+
+#include <algorithm>
 
 namespace arm_compute
 {
@@ -36,43 +34,6 @@
 {
 namespace neon
 {
-/** Helper to create an empty tensor.
- *
- * @param[in] shape                Desired shape.
- * @param[in] data_type            Desired data type.
- * @param[in] num_channels         (Optional) It indicates the number of channels for each tensor element
- * @param[in] fixed_point_position (Optional) Fixed point position that expresses the number of bits for the fractional part of the number when the tensor's data type is QS8 or QS16.
- *
- * @return Empty @ref Tensor with the specified shape and data type.
- */
-inline Tensor create_tensor(const TensorShape &shape, DataType data_type, int num_channels = 1, int fixed_point_position = 0)
-{
-    Tensor tensor;
-    tensor.allocator()->init(TensorInfo(shape, num_channels, data_type, fixed_point_position));
-
-    return tensor;
-}
-
-/** Helper to create an empty tensor.
- *
- * @param[in] name                 File name from which to get the dimensions.
- * @param[in] data_type            Desired data type.
- * @param[in] fixed_point_position (Optional) Number of bits for the fractional part of the fixed point numbers
- *
- * @return Empty @ref Tensor with the specified shape and data type.
- */
-inline Tensor create_tensor(const std::string &name, DataType data_type, int fixed_point_position = 0)
-{
-    constexpr unsigned int num_channels = 1;
-
-    const RawTensor &raw = library->get(name);
-
-    Tensor tensor;
-    tensor.allocator()->init(TensorInfo(raw.shape(), num_channels, data_type, fixed_point_position));
-
-    return tensor;
-}
-
 template <typename T>
 Array<T> create_array(const std::vector<T> &v)
 {
diff --git a/tests/Utils.h b/tests/Utils.h
index 3be0461..8d5046d 100644
--- a/tests/Utils.h
+++ b/tests/Utils.h
@@ -397,6 +397,24 @@
     }
     return true;
 }
+
+/** Create and initialize a tensor of the given type.
+ *
+ * @param[in] shape                Tensor shape.
+ * @param[in] data_type            Data type.
+ * @param[in] num_channels         (Optional) Number of channels.
+ * @param[in] fixed_point_position (Optional) Number of fractional bits.
+ *
+ * @return Initialized tensor of given type.
+ */
+template <typename T>
+inline T create_tensor(const TensorShape &shape, DataType data_type, int num_channels = 1, int fixed_point_position = 0)
+{
+    T tensor;
+    tensor.allocator()->init(TensorInfo(shape, num_channels, data_type, fixed_point_position));
+
+    return tensor;
+}
 } // namespace test
 } // namespace arm_compute
 #endif
diff --git a/tests/benchmark/CL/ActivationLayer.cpp b/tests/benchmark/CL/ActivationLayer.cpp
index 5180d3d..52a357b 100644
--- a/tests/benchmark/CL/ActivationLayer.cpp
+++ b/tests/benchmark/CL/ActivationLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/CL/BitwiseAnd.cpp b/tests/benchmark/CL/BitwiseAnd.cpp
index a3deb3e..4858c73 100644
--- a/tests/benchmark/CL/BitwiseAnd.cpp
+++ b/tests/benchmark/CL/BitwiseAnd.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
@@ -59,11 +58,12 @@
         profiler.add(std::make_shared<WallClockTimer>());
 
         const std::string image_name = *(DataSet().begin() + state.range(0));
+        const RawTensor &raw        = library->get(image_name);
 
         // Create tensors
-        src1 = create_tensor(image_name, DataType::U8);
-        src2 = create_tensor(image_name, DataType::U8);
-        dst  = create_tensor(image_name, DataType::U8);
+        src1 = create_tensor<CLTensor>(raw.shape(), DataType::U8);
+        src2 = create_tensor<CLTensor>(raw.shape(), DataType::U8);
+        dst  = create_tensor<CLTensor>(raw.shape(), DataType::U8);
 
         // Create and configure function
         band.configure(&src1, &src2, &dst);
diff --git a/tests/benchmark/CL/ConvolutionLayer.cpp b/tests/benchmark/CL/ConvolutionLayer.cpp
index e1f4fab..e790273 100644
--- a/tests/benchmark/CL/ConvolutionLayer.cpp
+++ b/tests/benchmark/CL/ConvolutionLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/CL/FullyConnectedLayer.cpp b/tests/benchmark/CL/FullyConnectedLayer.cpp
index 6e8c89f..fb8e1bc 100644
--- a/tests/benchmark/CL/FullyConnectedLayer.cpp
+++ b/tests/benchmark/CL/FullyConnectedLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/CL/GEMM.cpp b/tests/benchmark/CL/GEMM.cpp
index a10da09..87dad05 100644
--- a/tests/benchmark/CL/GEMM.cpp
+++ b/tests/benchmark/CL/GEMM.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/CL/GEMM.h b/tests/benchmark/CL/GEMM.h
index 02a3396..ca3d9ad 100644
--- a/tests/benchmark/CL/GEMM.h
+++ b/tests/benchmark/CL/GEMM.h
@@ -59,10 +59,10 @@
         TensorShape shape_d = gemm_obj.shape_d;
 
         // Create tensors
-        a = create_tensor(shape_a, data_type);
-        b = create_tensor(shape_b, data_type);
-        c = create_tensor(shape_c, data_type);
-        d = create_tensor(shape_d, data_type);
+        a = create_tensor<CLTensor>(shape_a, data_type);
+        b = create_tensor<CLTensor>(shape_b, data_type);
+        c = create_tensor<CLTensor>(shape_c, data_type);
+        d = create_tensor<CLTensor>(shape_d, data_type);
 
         // Create and configure function
         gemm_layer = std::unique_ptr<Function>(new Function());
diff --git a/tests/benchmark/CL/NormalizationLayer.cpp b/tests/benchmark/CL/NormalizationLayer.cpp
index 81d3c65..28f89dc 100644
--- a/tests/benchmark/CL/NormalizationLayer.cpp
+++ b/tests/benchmark/CL/NormalizationLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/CL/PoolingLayer.cpp b/tests/benchmark/CL/PoolingLayer.cpp
index 5285f27..05764a4 100644
--- a/tests/benchmark/CL/PoolingLayer.cpp
+++ b/tests/benchmark/CL/PoolingLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/NEON/ActivationLayer.cpp b/tests/benchmark/NEON/ActivationLayer.cpp
index 8faed9f..9ea675e 100644
--- a/tests/benchmark/NEON/ActivationLayer.cpp
+++ b/tests/benchmark/NEON/ActivationLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/NEON/BitwiseAnd.cpp b/tests/benchmark/NEON/BitwiseAnd.cpp
index dba3d1e..be68fd3 100644
--- a/tests/benchmark/NEON/BitwiseAnd.cpp
+++ b/tests/benchmark/NEON/BitwiseAnd.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
@@ -56,11 +55,12 @@
         profiler.add(std::make_shared<WallClockTimer>());
 
         const std::string image_name = *(DataSet().begin() + state.range(0));
+        const RawTensor &raw        = library->get(image_name);
 
         // Create tensors
-        src1 = create_tensor(image_name, DataType::U8);
-        src2 = create_tensor(image_name, DataType::U8);
-        dst  = create_tensor(image_name, DataType::U8);
+        src1 = create_tensor<Tensor>(raw.shape(), DataType::U8);
+        src2 = create_tensor<Tensor>(raw.shape(), DataType::U8);
+        dst  = create_tensor<Tensor>(raw.shape(), DataType::U8);
 
         // Create and configure function
         band.configure(&src1, &src2, &dst);
diff --git a/tests/benchmark/NEON/ConvolutionLayer.cpp b/tests/benchmark/NEON/ConvolutionLayer.cpp
index 0cfff84..a0b1236 100644
--- a/tests/benchmark/NEON/ConvolutionLayer.cpp
+++ b/tests/benchmark/NEON/ConvolutionLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/NEON/ConvolutionLayerDirect.cpp b/tests/benchmark/NEON/ConvolutionLayerDirect.cpp
index bc56e84..bbe5707 100644
--- a/tests/benchmark/NEON/ConvolutionLayerDirect.cpp
+++ b/tests/benchmark/NEON/ConvolutionLayerDirect.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/NEON/FullyConnectedLayer.cpp b/tests/benchmark/NEON/FullyConnectedLayer.cpp
index 8597920..15c2a40 100644
--- a/tests/benchmark/NEON/FullyConnectedLayer.cpp
+++ b/tests/benchmark/NEON/FullyConnectedLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/NEON/GEMM.cpp b/tests/benchmark/NEON/GEMM.cpp
index 50131b4..1655c32 100644
--- a/tests/benchmark/NEON/GEMM.cpp
+++ b/tests/benchmark/NEON/GEMM.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/NEON/GEMM.h b/tests/benchmark/NEON/GEMM.h
index 24d1965..4f0557d 100644
--- a/tests/benchmark/NEON/GEMM.h
+++ b/tests/benchmark/NEON/GEMM.h
@@ -63,10 +63,10 @@
         TensorShape shape_d = gemm_obj.shape_d;
 
         // Create tensors
-        a = create_tensor(shape_a, data_type, 1, 4);
-        b = create_tensor(shape_b, data_type, 1, 4);
-        c = create_tensor(shape_c, data_type, 1, 4);
-        d = create_tensor(shape_d, data_type, 1, 4);
+        a = create_tensor<Tensor>(shape_a, data_type, 1, 4);
+        b = create_tensor<Tensor>(shape_b, data_type, 1, 4);
+        c = create_tensor<Tensor>(shape_c, data_type, 1, 4);
+        d = create_tensor<Tensor>(shape_d, data_type, 1, 4);
 
         // Create and configure function
         gemm_layer = std::unique_ptr<Function>(new Function());
diff --git a/tests/benchmark/NEON/NormalizationLayer.cpp b/tests/benchmark/NEON/NormalizationLayer.cpp
index 46dc56b..8ea8c95c 100644
--- a/tests/benchmark/NEON/NormalizationLayer.cpp
+++ b/tests/benchmark/NEON/NormalizationLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/NEON/PoolingLayer.cpp b/tests/benchmark/NEON/PoolingLayer.cpp
index 9b07131..dfa7549 100644
--- a/tests/benchmark/NEON/PoolingLayer.cpp
+++ b/tests/benchmark/NEON/PoolingLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/common/ActivationLayer.h b/tests/benchmark/common/ActivationLayer.h
index 7edfb6e..8e22281 100644
--- a/tests/benchmark/common/ActivationLayer.h
+++ b/tests/benchmark/common/ActivationLayer.h
@@ -57,8 +57,8 @@
         shape.set(shape.num_dimensions(), batches);
 
         // Create tensors
-        src = create_tensor(shape, dt, 1, fixed_point_position);
-        dst = create_tensor(shape, dt, 1, fixed_point_position);
+        src = create_tensor<TensorType>(shape, dt, 1, fixed_point_position);
+        dst = create_tensor<TensorType>(shape, dt, 1, fixed_point_position);
 
         // Create and configure function
         act_layer.configure(&src, &dst, act_obj.info);
diff --git a/tests/benchmark/common/ConvolutionLayer.h b/tests/benchmark/common/ConvolutionLayer.h
index 594c62c..a777a95 100644
--- a/tests/benchmark/common/ConvolutionLayer.h
+++ b/tests/benchmark/common/ConvolutionLayer.h
@@ -59,10 +59,10 @@
         dst_shape.set(3 /* batch */, batches);
 
         // Create tensors
-        src     = create_tensor(src_shape, dt, 1, fixed_point_position);
-        weights = create_tensor(conv_obj.weights_shape, dt, 1, fixed_point_position);
-        bias    = create_tensor(conv_obj.bias_shape, dt, 1, fixed_point_position);
-        dst     = create_tensor(dst_shape, dt, 1, fixed_point_position);
+        src     = create_tensor<TensorType>(src_shape, dt, 1, fixed_point_position);
+        weights = create_tensor<TensorType>(conv_obj.weights_shape, dt, 1, fixed_point_position);
+        bias    = create_tensor<TensorType>(conv_obj.bias_shape, dt, 1, fixed_point_position);
+        dst     = create_tensor<TensorType>(dst_shape, dt, 1, fixed_point_position);
 
         // Create and configure function
         conv_layer = std::unique_ptr<Function>(new Function());
diff --git a/tests/benchmark/common/FullyConnectedLayer.h b/tests/benchmark/common/FullyConnectedLayer.h
index 88adf83..eeef1de 100644
--- a/tests/benchmark/common/FullyConnectedLayer.h
+++ b/tests/benchmark/common/FullyConnectedLayer.h
@@ -60,10 +60,10 @@
         dst_shape.set(dst_shape.num_dimensions(), batches);
 
         // Create tensors
-        src     = create_tensor(src_shape, dt, 1, fixed_point_position);
-        weights = create_tensor(fc_obj.weights_shape, dt, 1, fixed_point_position);
-        bias    = create_tensor(fc_obj.bias_shape, dt, 1, fixed_point_position);
-        dst     = create_tensor(dst_shape, dt, 1, fixed_point_position);
+        src     = create_tensor<TensorType>(src_shape, dt, 1, fixed_point_position);
+        weights = create_tensor<TensorType>(fc_obj.weights_shape, dt, 1, fixed_point_position);
+        bias    = create_tensor<TensorType>(fc_obj.bias_shape, dt, 1, fixed_point_position);
+        dst     = create_tensor<TensorType>(dst_shape, dt, 1, fixed_point_position);
 
         // Create and configure function
         fc_layer = std::unique_ptr<Function>(new Function());
diff --git a/tests/benchmark/common/NormalizationLayer.h b/tests/benchmark/common/NormalizationLayer.h
index 4593fb7..580dca6 100644
--- a/tests/benchmark/common/NormalizationLayer.h
+++ b/tests/benchmark/common/NormalizationLayer.h
@@ -58,8 +58,8 @@
         shape.set(shape.num_dimensions(), batches);
 
         // Create tensors
-        src = create_tensor(shape, dt, 1, fixed_point_position);
-        dst = create_tensor(shape, dt, 1, fixed_point_position);
+        src = create_tensor<TensorType>(shape, dt, 1, fixed_point_position);
+        dst = create_tensor<TensorType>(shape, dt, 1, fixed_point_position);
 
         // Create and configure function
         norm_layer = std::unique_ptr<Function>(new Function());
diff --git a/tests/benchmark/common/PoolingLayer.h b/tests/benchmark/common/PoolingLayer.h
index 5bb332f..96ff905 100644
--- a/tests/benchmark/common/PoolingLayer.h
+++ b/tests/benchmark/common/PoolingLayer.h
@@ -59,8 +59,8 @@
         dst_shape.set(dst_shape.num_dimensions(), batches);
 
         // Create tensors
-        src = create_tensor(src_shape, dt, 1, fixed_point_position);
-        dst = create_tensor(dst_shape, dt, 1, fixed_point_position);
+        src = create_tensor<TensorType>(src_shape, dt, 1, fixed_point_position);
+        dst = create_tensor<TensorType>(dst_shape, dt, 1, fixed_point_position);
 
         // Create and configure function
         pool_layer.configure(&src, &dst, pool_obj.info);
diff --git a/tests/benchmark/system_tests/CL/AlexNet.cpp b/tests/benchmark/system_tests/CL/AlexNet.cpp
index fe0b991..04b65b8 100644
--- a/tests/benchmark/system_tests/CL/AlexNet.cpp
+++ b/tests/benchmark/system_tests/CL/AlexNet.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/system_tests/CL/LeNet5.cpp b/tests/benchmark/system_tests/CL/LeNet5.cpp
index d65a7dd..60bb37a 100644
--- a/tests/benchmark/system_tests/CL/LeNet5.cpp
+++ b/tests/benchmark/system_tests/CL/LeNet5.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/system_tests/NEON/AlexNet.cpp b/tests/benchmark/system_tests/NEON/AlexNet.cpp
index 2d222e7..530c400 100644
--- a/tests/benchmark/system_tests/NEON/AlexNet.cpp
+++ b/tests/benchmark/system_tests/NEON/AlexNet.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/system_tests/NEON/LeNet5.cpp b/tests/benchmark/system_tests/NEON/LeNet5.cpp
index 5170f05..8724d8b 100644
--- a/tests/benchmark/system_tests/NEON/LeNet5.cpp
+++ b/tests/benchmark/system_tests/NEON/LeNet5.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "benchmark/Datasets.h"
diff --git a/tests/benchmark/system_tests/common/AlexNet.h b/tests/benchmark/system_tests/common/AlexNet.h
index 9c93dc7..46e26bf 100644
--- a/tests/benchmark/system_tests/common/AlexNet.h
+++ b/tests/benchmark/system_tests/common/AlexNet.h
@@ -24,8 +24,8 @@
 #ifndef __ARM_COMPUTE_TEST_BENCHMARK_ALEXNET_H__
 #define __ARM_COMPUTE_TEST_BENCHMARK_ALEXNET_H__
 
-#include "TensorLibrary.h"
-#include "Utils.h"
+#include "tests/TensorLibrary.h"
+#include "tests/Utils.h"
 
 #include "benchmark/Profiler.h"
 #include "benchmark/WallClockTimer.h"
diff --git a/tests/benchmark/system_tests/common/LeNet5.h b/tests/benchmark/system_tests/common/LeNet5.h
index db34f68..cca1bbf 100644
--- a/tests/benchmark/system_tests/common/LeNet5.h
+++ b/tests/benchmark/system_tests/common/LeNet5.h
@@ -24,8 +24,8 @@
 #ifndef __ARM_COMPUTE_TEST_BENCHMARK_LENET5_H__
 #define __ARM_COMPUTE_TEST_BENCHMARK_LENET5_H__
 
-#include "TensorLibrary.h"
-#include "Utils.h"
+#include "tests/TensorLibrary.h"
+#include "tests/Utils.h"
 
 #include "benchmark/Profiler.h"
 #include "benchmark/WallClockTimer.h"
@@ -79,4 +79,4 @@
 } // namespace benchmark
 } // namespace test
 } // namespace arm_compute
-#endif //__ARM_COMPUTE_TEST_BENCHMARK_LENET5_H__
+#endif
diff --git a/tests/model_objects/AlexNet.h b/tests/model_objects/AlexNet.h
index 8443e9c..672942d 100644
--- a/tests/model_objects/AlexNet.h
+++ b/tests/model_objects/AlexNet.h
@@ -24,8 +24,9 @@
 #ifndef __ARM_COMPUTE_TEST_MODEL_OBJECTS_ALEXNET_H__
 #define __ARM_COMPUTE_TEST_MODEL_OBJECTS_ALEXNET_H__
 
-#include "TensorLibrary.h"
-#include "Utils.h"
+#include "tests/Globals.h"
+#include "tests/TensorLibrary.h"
+#include "tests/Utils.h"
 
 #include <memory>
 
diff --git a/tests/model_objects/LeNet5.h b/tests/model_objects/LeNet5.h
index 7d5090f..3a9046c 100644
--- a/tests/model_objects/LeNet5.h
+++ b/tests/model_objects/LeNet5.h
@@ -24,8 +24,9 @@
 #ifndef __ARM_COMPUTE_TEST_MODEL_OBJECTS_LENET5_H__
 #define __ARM_COMPUTE_TEST_MODEL_OBJECTS_LENET5_H__
 
-#include "TensorLibrary.h"
-#include "Utils.h"
+#include "tests/Globals.h"
+#include "tests/TensorLibrary.h"
+#include "tests/Utils.h"
 
 #include <memory>
 
diff --git a/tests/validation/CL/ActivationLayer.cpp b/tests/validation/CL/ActivationLayer.cpp
index 793ab30..5d266e6 100644
--- a/tests/validation/CL/ActivationLayer.cpp
+++ b/tests/validation/CL/ActivationLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -87,8 +86,8 @@
 CLTensor compute_activation_layer(bool in_place, const TensorShape &shape, DataType dt, ActivationLayerInfo act_info, int fixed_point_position = 0)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, dt, 1, fixed_point_position);
-    CLTensor dst = create_tensor(shape, dt, 1, fixed_point_position);
+    CLTensor src = create_tensor<CLTensor>(shape, dt, 1, fixed_point_position);
+    CLTensor dst = create_tensor<CLTensor>(shape, dt, 1, fixed_point_position);
 
     // Create and configure function
     CLActivationLayer act_layer;
@@ -155,8 +154,8 @@
     const int fixed_point_position = (arm_compute::is_data_type_fixed_point(dt)) ? 3 : 0;
 
     // Create tensors
-    CLTensor src = create_tensor(shape, dt, 1, fixed_point_position);
-    CLTensor dst = create_tensor(shape, dt, 1, fixed_point_position);
+    CLTensor src = create_tensor<CLTensor>(shape, dt, 1, fixed_point_position);
+    CLTensor dst = create_tensor<CLTensor>(shape, dt, 1, fixed_point_position);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/CL/BitwiseAnd.cpp b/tests/validation/CL/BitwiseAnd.cpp
index 8b40aff..c818fc5 100644
--- a/tests/validation/CL/BitwiseAnd.cpp
+++ b/tests/validation/CL/BitwiseAnd.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -60,9 +59,9 @@
 CLTensor compute_bitwise_and(const TensorShape &shape)
 {
     // Create tensors
-    CLTensor src1 = create_tensor(shape, DataType::U8);
-    CLTensor src2 = create_tensor(shape, DataType::U8);
-    CLTensor dst  = create_tensor(shape, DataType::U8);
+    CLTensor src1 = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor src2 = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst  = create_tensor<CLTensor>(shape, DataType::U8);
 
     // Create and configure function
     CLBitwiseAnd band;
@@ -96,9 +95,9 @@
 CLTensor compute_bitwise_and_subtensor(const TensorShape &shape)
 {
     // Create tensors
-    CLTensor src1 = create_tensor(shape, DataType::U8);
-    CLTensor src2 = create_tensor(shape, DataType::U8);
-    CLTensor dst  = create_tensor(shape, DataType::U8);
+    CLTensor src1 = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor src2 = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst  = create_tensor<CLTensor>(shape, DataType::U8);
 
     // Create SubTensors
     int         coord_z   = shape.z() / 2;
@@ -147,9 +146,9 @@
 BOOST_DATA_TEST_CASE(Configuration, SmallShapes() + LargeShapes(), shape)
 {
     // Create tensors
-    CLTensor src1 = create_tensor(shape, DataType::U8);
-    CLTensor src2 = create_tensor(shape, DataType::U8);
-    CLTensor dst  = create_tensor(shape, DataType::U8);
+    CLTensor src1 = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor src2 = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst  = create_tensor<CLTensor>(shape, DataType::U8);
 
     BOOST_TEST(src1.info()->is_resizable());
     BOOST_TEST(src2.info()->is_resizable());
diff --git a/tests/validation/CL/Box3x3.cpp b/tests/validation/CL/Box3x3.cpp
index 9c3db9f..96a0d09 100644
--- a/tests/validation/CL/Box3x3.cpp
+++ b/tests/validation/CL/Box3x3.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -65,8 +64,8 @@
 CLTensor compute_box3x3(const TensorShape &shape, BorderMode border_mode, uint8_t constant_border_value)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U8);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U8);
 
     // Create and configure function
     CLBox3x3 box3x3;
@@ -97,8 +96,8 @@
 BOOST_DATA_TEST_CASE(Configuration, (SmallShapes() + LargeShapes()) * BorderModes(), shape, border_mode)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U8);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/CL/CMakeLists.txt b/tests/validation/CL/CMakeLists.txt
index d46cca7..93547b1 100644
--- a/tests/validation/CL/CMakeLists.txt
+++ b/tests/validation/CL/CMakeLists.txt
@@ -25,7 +25,6 @@
 
 set(arm_compute_test_validation_OPENCL_SOURCE_FILES
     ${CMAKE_SOURCE_DIR}/CL/CLAccessor.h
-    ${CMAKE_SOURCE_DIR}/CL/Helper.h
     ${CMAKE_CURRENT_SOURCE_DIR}/CLFixture.h
     ${CMAKE_CURRENT_SOURCE_DIR}/CLFixture.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/BitwiseAnd.cpp
diff --git a/tests/validation/CL/DepthConvert.cpp b/tests/validation/CL/DepthConvert.cpp
index 0bca453..87fe381 100644
--- a/tests/validation/CL/DepthConvert.cpp
+++ b/tests/validation/CL/DepthConvert.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -63,8 +62,8 @@
 CLTensor compute_depth_convert(const TensorShape &shape, DataType dt_in, DataType dt_out, ConvertPolicy policy, uint32_t shift)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, dt_in);
-    CLTensor dst = create_tensor(shape, dt_out);
+    CLTensor src = create_tensor<CLTensor>(shape, dt_in);
+    CLTensor dst = create_tensor<CLTensor>(shape, dt_out);
 
     // Create and configure function
     CLDepthConvert depth_convert;
@@ -97,8 +96,8 @@
 void compute_configure_validate(const TensorShape &shape, DataType dt_in, DataType dt_out, ConvertPolicy policy, uint32_t shift)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, dt_in);
-    CLTensor dst = create_tensor(shape, dt_out);
+    CLTensor src = create_tensor<CLTensor>(shape, dt_in);
+    CLTensor dst = create_tensor<CLTensor>(shape, dt_out);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/CL/FillBorder.cpp b/tests/validation/CL/FillBorder.cpp
index 42b9064..f6930ee 100644
--- a/tests/validation/CL/FillBorder.cpp
+++ b/tests/validation/CL/FillBorder.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -58,7 +57,7 @@
     BorderSize        border_size{ 5 };
 
     // Create tensors
-    CLTensor src = create_tensor(TensorShape{ 10U, 10U, 2U }, DataType::U8);
+    CLTensor src = create_tensor<CLTensor>(TensorShape{ 10U, 10U, 2U }, DataType::U8);
 
     src.info()->extend_padding(padding);
 
diff --git a/tests/validation/CL/Gaussian3x3.cpp b/tests/validation/CL/Gaussian3x3.cpp
index 2a526e8..bf4684d 100644
--- a/tests/validation/CL/Gaussian3x3.cpp
+++ b/tests/validation/CL/Gaussian3x3.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -65,8 +64,8 @@
 CLTensor compute_gaussian3x3(const TensorShape &shape, BorderMode border_mode, uint8_t constant_border_value)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U8);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U8);
 
     // Create and configure function
     CLGaussian3x3 gaussian3x3;
@@ -97,8 +96,8 @@
 BOOST_DATA_TEST_CASE(Configuration, (SmallShapes() + LargeShapes()) * BorderModes(), shape, border_mode)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U8);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/CL/Gaussian5x5.cpp b/tests/validation/CL/Gaussian5x5.cpp
index b6a692c..827dadd 100644
--- a/tests/validation/CL/Gaussian5x5.cpp
+++ b/tests/validation/CL/Gaussian5x5.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -65,8 +64,8 @@
 CLTensor compute_gaussian5x5(const TensorShape &shape, BorderMode border_mode, uint8_t constant_border_value)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U8);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U8);
 
     // Create and configure function
     CLGaussian5x5 gaussian5x5;
@@ -97,8 +96,8 @@
 BOOST_DATA_TEST_CASE(Configuration, (SmallShapes() + LargeShapes()) * BorderModes(), shape, border_mode)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U8);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/CL/IntegralImage.cpp b/tests/validation/CL/IntegralImage.cpp
index 9ace8bd..e46568e 100644
--- a/tests/validation/CL/IntegralImage.cpp
+++ b/tests/validation/CL/IntegralImage.cpp
@@ -23,7 +23,6 @@
  */
 
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -60,8 +59,8 @@
 CLTensor compute_integral_image(const TensorShape &shape)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U32);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U32);
 
     // Create integral image configure function
     CLIntegralImage integral_image;
@@ -92,8 +91,8 @@
 BOOST_DATA_TEST_CASE(Configuration, SmallShapes() + LargeShapes(), shape)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U32);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U32);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/CL/NonLinearFilter.cpp b/tests/validation/CL/NonLinearFilter.cpp
index 4a487d6..414a308 100644
--- a/tests/validation/CL/NonLinearFilter.cpp
+++ b/tests/validation/CL/NonLinearFilter.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -69,8 +68,8 @@
                                    uint8_t constant_border_value)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U8);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U8);
 
     // Create and configure function
     CLNonLinearFilter filter;
@@ -113,8 +112,8 @@
     const auto half_mask_size = static_cast<int>(mask_size / 2);
 
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U8);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/CL/PoolingLayer.cpp b/tests/validation/CL/PoolingLayer.cpp
index dabc21e..5876e8b 100644
--- a/tests/validation/CL/PoolingLayer.cpp
+++ b/tests/validation/CL/PoolingLayer.cpp
@@ -22,9 +22,10 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "TypePrinter.h"
 #include "arm_compute/runtime/CL/functions/CLPoolingLayer.h"
+#include "tests/Globals.h"
+#include "tests/Utils.h"
 #include "tests/dataset/PoolingLayerDataset.h"
 #include "validation/Datasets.h"
 #include "validation/Reference.h"
@@ -52,8 +53,8 @@
 CLTensor compute_pooling_layer(const TensorShape &shape_in, const TensorShape &shape_out, DataType dt, PoolingLayerInfo pool_info)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape_in, dt);
-    CLTensor dst = create_tensor(shape_out, dt);
+    CLTensor src = create_tensor<CLTensor>(shape_in, dt);
+    CLTensor dst = create_tensor<CLTensor>(shape_out, dt);
 
     // Create and configure function
     CLPoolingLayer pool;
diff --git a/tests/validation/CL/Sobel3x3.cpp b/tests/validation/CL/Sobel3x3.cpp
index fbd7185..59c21e7 100644
--- a/tests/validation/CL/Sobel3x3.cpp
+++ b/tests/validation/CL/Sobel3x3.cpp
@@ -23,7 +23,6 @@
  */
 
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -67,9 +66,9 @@
 std::pair<CLTensor, CLTensor> compute_sobel_3x3(const TensorShape &shape, BorderMode border_mode, uint8_t constant_border_value)
 {
     // Create tensors
-    CLTensor src   = create_tensor(shape, DataType::U8);
-    CLTensor dst_x = create_tensor(shape, DataType::S16);
-    CLTensor dst_y = create_tensor(shape, DataType::S16);
+    CLTensor src   = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst_x = create_tensor<CLTensor>(shape, DataType::S16);
+    CLTensor dst_y = create_tensor<CLTensor>(shape, DataType::S16);
 
     src.info()->set_format(Format::U8);
     dst_x.info()->set_format(Format::S16);
@@ -106,9 +105,9 @@
 BOOST_DATA_TEST_CASE(Configuration, (SmallShapes() + LargeShapes()) * BorderModes(), shape, border_mode)
 {
     // Create tensors
-    CLTensor src   = create_tensor(shape, DataType::U8);
-    CLTensor dst_x = create_tensor(shape, DataType::S16);
-    CLTensor dst_y = create_tensor(shape, DataType::S16);
+    CLTensor src   = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst_x = create_tensor<CLTensor>(shape, DataType::S16);
+    CLTensor dst_y = create_tensor<CLTensor>(shape, DataType::S16);
 
     src.info()->set_format(Format::U8);
     dst_x.info()->set_format(Format::S16);
diff --git a/tests/validation/CL/Sobel5x5.cpp b/tests/validation/CL/Sobel5x5.cpp
index 263e57e..8beb795 100644
--- a/tests/validation/CL/Sobel5x5.cpp
+++ b/tests/validation/CL/Sobel5x5.cpp
@@ -23,7 +23,6 @@
  */
 
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -67,9 +66,9 @@
 std::pair<CLTensor, CLTensor> compute_sobel_5x5(const TensorShape &shape, BorderMode border_mode, uint8_t constant_border_value)
 {
     // Create tensors
-    CLTensor src   = create_tensor(shape, DataType::U8);
-    CLTensor dst_x = create_tensor(shape, DataType::S16);
-    CLTensor dst_y = create_tensor(shape, DataType::S16);
+    CLTensor src   = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst_x = create_tensor<CLTensor>(shape, DataType::S16);
+    CLTensor dst_y = create_tensor<CLTensor>(shape, DataType::S16);
 
     src.info()->set_format(Format::U8);
     dst_x.info()->set_format(Format::S16);
@@ -106,9 +105,9 @@
 BOOST_DATA_TEST_CASE(Configuration, (SmallShapes() + LargeShapes()) * BorderModes(), shape, border_mode)
 {
     // Create tensors
-    CLTensor src   = create_tensor(shape, DataType::U8);
-    CLTensor dst_x = create_tensor(shape, DataType::S16);
-    CLTensor dst_y = create_tensor(shape, DataType::S16);
+    CLTensor src   = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst_x = create_tensor<CLTensor>(shape, DataType::S16);
+    CLTensor dst_y = create_tensor<CLTensor>(shape, DataType::S16);
 
     src.info()->set_format(Format::U8);
     dst_x.info()->set_format(Format::S16);
diff --git a/tests/validation/CL/SoftmaxLayer.cpp b/tests/validation/CL/SoftmaxLayer.cpp
index aa80e50..3f8f465 100644
--- a/tests/validation/CL/SoftmaxLayer.cpp
+++ b/tests/validation/CL/SoftmaxLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -66,8 +65,8 @@
 CLTensor compute_softmax_layer(const TensorShape &shape, DataType dt, int fixed_point_position = 0)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, dt, 1, fixed_point_position);
-    CLTensor dst = create_tensor(shape, dt, 1, fixed_point_position);
+    CLTensor src = create_tensor<CLTensor>(shape, dt, 1, fixed_point_position);
+    CLTensor dst = create_tensor<CLTensor>(shape, dt, 1, fixed_point_position);
 
     // Create and configure function
     CLSoftmaxLayer smx_layer;
@@ -111,8 +110,8 @@
     int fixed_point_position = (arm_compute::is_data_type_fixed_point(dt)) ? 3 : 0;
 
     // Create tensors
-    CLTensor src = create_tensor(shape, dt, 1, fixed_point_position);
-    CLTensor dst = create_tensor(shape, dt, 1, fixed_point_position);
+    CLTensor src = create_tensor<CLTensor>(shape, dt, 1, fixed_point_position);
+    CLTensor dst = create_tensor<CLTensor>(shape, dt, 1, fixed_point_position);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/CL/Threshold.cpp b/tests/validation/CL/Threshold.cpp
index ff7b8f0..4f9e420 100644
--- a/tests/validation/CL/Threshold.cpp
+++ b/tests/validation/CL/Threshold.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "Globals.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -65,8 +64,8 @@
 CLTensor compute_threshold(const TensorShape &shape, uint8_t threshold, uint8_t false_value, uint8_t true_value, ThresholdType type, uint8_t upper)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U8);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U8);
 
     // Create and configure function
     CLThreshold thrsh;
@@ -99,8 +98,8 @@
                      shape, threshold_conf)
 {
     // Create tensors
-    CLTensor src = create_tensor(shape, DataType::U8);
-    CLTensor dst = create_tensor(shape, DataType::U8);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/AbsoluteDifference.cpp b/tests/validation/NEON/AbsoluteDifference.cpp
index 334a842..779baa7 100644
--- a/tests/validation/NEON/AbsoluteDifference.cpp
+++ b/tests/validation/NEON/AbsoluteDifference.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -62,9 +61,9 @@
 Tensor compute_absolute_difference(const TensorShape &shape, DataType dt_in0, DataType dt_in1, DataType dt_out)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, dt_in0);
-    Tensor src2 = create_tensor(shape, dt_in1);
-    Tensor dst  = create_tensor(shape, dt_out);
+    Tensor src1 = create_tensor<Tensor>(shape, dt_in0);
+    Tensor src2 = create_tensor<Tensor>(shape, dt_in1);
+    Tensor dst  = create_tensor<Tensor>(shape, dt_out);
 
     // Create and configure function
     NEAbsoluteDifference abs_d;
@@ -123,9 +122,9 @@
                      shape)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     validate_configuration(src1, src2, dst, shape);
 }
@@ -163,9 +162,9 @@
                      shape, dt)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, dt);
-    Tensor src2 = create_tensor(shape, DataType::S16);
-    Tensor dst  = create_tensor(shape, DataType::S16);
+    Tensor src1 = create_tensor<Tensor>(shape, dt);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::S16);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::S16);
 
     validate_configuration(src1, src2, dst, shape);
 }
diff --git a/tests/validation/NEON/Accumulate.cpp b/tests/validation/NEON/Accumulate.cpp
index 157c298..f8cc507 100644
--- a/tests/validation/NEON/Accumulate.cpp
+++ b/tests/validation/NEON/Accumulate.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -59,8 +58,8 @@
 Tensor compute_accumulate(const TensorShape &shape)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::S16);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::S16);
 
     // Create and configure function
     NEAccumulate acc;
@@ -93,8 +92,8 @@
                      shape)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::S16);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::S16);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/AccumulateSquared.cpp b/tests/validation/NEON/AccumulateSquared.cpp
index 3dbdeb2..093ccb4 100644
--- a/tests/validation/NEON/AccumulateSquared.cpp
+++ b/tests/validation/NEON/AccumulateSquared.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -59,8 +58,8 @@
 Tensor compute_accumulate_squared(const TensorShape &shape, uint32_t shift)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::S16);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::S16);
 
     // Create and configure function
     NEAccumulateSquared acc;
@@ -94,8 +93,8 @@
                      shape, shift)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::S16);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::S16);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/AccumulateWeighted.cpp b/tests/validation/NEON/AccumulateWeighted.cpp
index cf1f746..8c54076 100644
--- a/tests/validation/NEON/AccumulateWeighted.cpp
+++ b/tests/validation/NEON/AccumulateWeighted.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -59,8 +58,8 @@
 Tensor compute_accumulate_weighted(const TensorShape &shape, float alpha)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create and configure function
     NEAccumulateWeighted acc;
@@ -93,8 +92,8 @@
                      shape, alpha)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/ActivationLayer.cpp b/tests/validation/NEON/ActivationLayer.cpp
index 11ac669..6a84a53 100644
--- a/tests/validation/NEON/ActivationLayer.cpp
+++ b/tests/validation/NEON/ActivationLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -87,8 +86,8 @@
 Tensor compute_activation_layer(bool in_place, const TensorShape &shape, DataType dt, ActivationLayerInfo act_info, int fixed_point_position = 0)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, dt, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, dt, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, dt, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, dt, 1, fixed_point_position);
 
     // Create and configure function
     NEActivationLayer act_layer;
@@ -155,8 +154,8 @@
     const int fixed_point_position = (arm_compute::is_data_type_fixed_point(dt)) ? 3 : 0;
 
     // Create tensors
-    Tensor src = create_tensor(shape, dt, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, dt, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, dt, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, dt, 1, fixed_point_position);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/ArithmeticAddition.cpp b/tests/validation/NEON/ArithmeticAddition.cpp
index d58d0ce..fe99829 100644
--- a/tests/validation/NEON/ArithmeticAddition.cpp
+++ b/tests/validation/NEON/ArithmeticAddition.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -63,9 +62,9 @@
 Tensor compute_arithmetic_addition(const TensorShape &shape, DataType dt_in0, DataType dt_in1, DataType dt_out, ConvertPolicy policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, dt_in0);
-    Tensor src2 = create_tensor(shape, dt_in1);
-    Tensor dst  = create_tensor(shape, dt_out);
+    Tensor src1 = create_tensor<Tensor>(shape, dt_in0);
+    Tensor src2 = create_tensor<Tensor>(shape, dt_in1);
+    Tensor dst  = create_tensor<Tensor>(shape, dt_out);
 
     // Create and configure function
     NEArithmeticAddition add;
@@ -124,9 +123,9 @@
                      shape, policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     validate_configuration(src1, src2, dst, shape, policy);
 }
@@ -151,9 +150,9 @@
                      shape, dt, policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, dt);
-    Tensor src2 = create_tensor(shape, DataType::S16);
-    Tensor dst  = create_tensor(shape, DataType::S16);
+    Tensor src1 = create_tensor<Tensor>(shape, dt);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::S16);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::S16);
 
     validate_configuration(src1, src2, dst, shape, policy);
 }
@@ -191,9 +190,9 @@
                      shape, policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::F32);
-    Tensor src2 = create_tensor(shape, DataType::F32);
-    Tensor dst  = create_tensor(shape, DataType::F32);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::F32);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::F32);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::F32);
 
     validate_configuration(src1, src2, dst, shape, policy);
 }
diff --git a/tests/validation/NEON/ArithmeticSubtraction.cpp b/tests/validation/NEON/ArithmeticSubtraction.cpp
index ade473a..35c2963 100644
--- a/tests/validation/NEON/ArithmeticSubtraction.cpp
+++ b/tests/validation/NEON/ArithmeticSubtraction.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -63,9 +62,9 @@
 Tensor compute_arithmetic_subtraction(const TensorShape &shape, DataType dt_in0, DataType dt_in1, DataType dt_out, ConvertPolicy policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, dt_in0);
-    Tensor src2 = create_tensor(shape, dt_in1);
-    Tensor dst  = create_tensor(shape, dt_out);
+    Tensor src1 = create_tensor<Tensor>(shape, dt_in0);
+    Tensor src2 = create_tensor<Tensor>(shape, dt_in1);
+    Tensor dst  = create_tensor<Tensor>(shape, dt_out);
 
     // Create and configure function
     NEArithmeticSubtraction sub;
@@ -124,9 +123,9 @@
                      shape, policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     validate_configuration(src1, src2, dst, shape, policy);
 }
@@ -151,9 +150,9 @@
                      shape, dt, policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, dt);
-    Tensor src2 = create_tensor(shape, DataType::S16);
-    Tensor dst  = create_tensor(shape, DataType::S16);
+    Tensor src1 = create_tensor<Tensor>(shape, dt);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::S16);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::S16);
 
     validate_configuration(src1, src2, dst, shape, policy);
 }
@@ -191,9 +190,9 @@
                      shape, policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::F32);
-    Tensor src2 = create_tensor(shape, DataType::F32);
-    Tensor dst  = create_tensor(shape, DataType::F32);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::F32);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::F32);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::F32);
 
     validate_configuration(src1, src2, dst, shape, policy);
 }
diff --git a/tests/validation/NEON/BatchNormalizationLayer.cpp b/tests/validation/NEON/BatchNormalizationLayer.cpp
index 7656b2f..3206b39 100644
--- a/tests/validation/NEON/BatchNormalizationLayer.cpp
+++ b/tests/validation/NEON/BatchNormalizationLayer.cpp
@@ -21,10 +21,11 @@
  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  */
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TypePrinter.h"
 #include "dataset/BatchNormalizationLayerDataset.h"
+#include "tests/Globals.h"
+#include "tests/Utils.h"
 #include "tests/validation/Helpers.h"
 #include "validation/Datasets.h"
 #include "validation/Reference.h"
@@ -55,12 +56,12 @@
 Tensor compute_reference_batch_normalization_layer(const TensorShape &shape0, const TensorShape &shape1, DataType dt, float epsilon, int fixed_point_position = 0)
 {
     // Create tensors
-    Tensor src   = create_tensor(shape0, dt, 1, fixed_point_position);
-    Tensor dst   = create_tensor(shape0, dt, 1, fixed_point_position);
-    Tensor mean  = create_tensor(shape1, dt, 1, fixed_point_position);
-    Tensor var   = create_tensor(shape1, dt, 1, fixed_point_position);
-    Tensor beta  = create_tensor(shape1, dt, 1, fixed_point_position);
-    Tensor gamma = create_tensor(shape1, dt, 1, fixed_point_position);
+    Tensor src   = create_tensor<Tensor>(shape0, dt, 1, fixed_point_position);
+    Tensor dst   = create_tensor<Tensor>(shape0, dt, 1, fixed_point_position);
+    Tensor mean  = create_tensor<Tensor>(shape1, dt, 1, fixed_point_position);
+    Tensor var   = create_tensor<Tensor>(shape1, dt, 1, fixed_point_position);
+    Tensor beta  = create_tensor<Tensor>(shape1, dt, 1, fixed_point_position);
+    Tensor gamma = create_tensor<Tensor>(shape1, dt, 1, fixed_point_position);
 
     // Create and configure function
     NEBatchNormalizationLayer norm;
@@ -127,12 +128,12 @@
     int fixed_point_position = (arm_compute::is_data_type_fixed_point(dt)) ? 3 : 0;
 
     // Create tensors
-    Tensor src   = create_tensor(obj.shape0, dt, 1, fixed_point_position);
-    Tensor dst   = create_tensor(obj.shape0, dt, 1, fixed_point_position);
-    Tensor mean  = create_tensor(obj.shape1, dt, 1, fixed_point_position);
-    Tensor var   = create_tensor(obj.shape1, dt, 1, fixed_point_position);
-    Tensor beta  = create_tensor(obj.shape1, dt, 1, fixed_point_position);
-    Tensor gamma = create_tensor(obj.shape1, dt, 1, fixed_point_position);
+    Tensor src   = create_tensor<Tensor>(obj.shape0, dt, 1, fixed_point_position);
+    Tensor dst   = create_tensor<Tensor>(obj.shape0, dt, 1, fixed_point_position);
+    Tensor mean  = create_tensor<Tensor>(obj.shape1, dt, 1, fixed_point_position);
+    Tensor var   = create_tensor<Tensor>(obj.shape1, dt, 1, fixed_point_position);
+    Tensor beta  = create_tensor<Tensor>(obj.shape1, dt, 1, fixed_point_position);
+    Tensor gamma = create_tensor<Tensor>(obj.shape1, dt, 1, fixed_point_position);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/BitwiseAnd.cpp b/tests/validation/NEON/BitwiseAnd.cpp
index 3dcad6f..44d0bd9 100644
--- a/tests/validation/NEON/BitwiseAnd.cpp
+++ b/tests/validation/NEON/BitwiseAnd.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -60,9 +59,9 @@
 Tensor compute_bitwise_and(const TensorShape &shape)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create and configure function
     NEBitwiseAnd band;
@@ -96,9 +95,9 @@
 Tensor compute_bitwise_and_subtensor(const TensorShape &shape)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create SubTensors
     int         coord_z   = shape.z() / 2;
@@ -147,9 +146,9 @@
 BOOST_DATA_TEST_CASE(Configuration, SmallShapes() + LargeShapes(), shape)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     BOOST_TEST(src1.info()->is_resizable());
     BOOST_TEST(src2.info()->is_resizable());
diff --git a/tests/validation/NEON/BitwiseNot.cpp b/tests/validation/NEON/BitwiseNot.cpp
index de58f55..d30339e 100644
--- a/tests/validation/NEON/BitwiseNot.cpp
+++ b/tests/validation/NEON/BitwiseNot.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -59,8 +58,8 @@
 Tensor compute_bitwise_not(const TensorShape &shape)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create not configure function
     NEBitwiseNot bnot;
@@ -91,8 +90,8 @@
 BOOST_DATA_TEST_CASE(Configuration, SmallShapes() + LargeShapes(), shape)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/BitwiseOr.cpp b/tests/validation/NEON/BitwiseOr.cpp
index 95f77ee..9b47f79 100644
--- a/tests/validation/NEON/BitwiseOr.cpp
+++ b/tests/validation/NEON/BitwiseOr.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -59,9 +58,9 @@
 Tensor compute_bitwise_or(const TensorShape &shape)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create and configure function
     NEBitwiseOr bor;
@@ -95,9 +94,9 @@
 BOOST_DATA_TEST_CASE(Configuration, SmallShapes() + LargeShapes(), shape)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     BOOST_TEST(src1.info()->is_resizable());
     BOOST_TEST(src2.info()->is_resizable());
diff --git a/tests/validation/NEON/BitwiseXor.cpp b/tests/validation/NEON/BitwiseXor.cpp
index a9225e6..7ab6448 100644
--- a/tests/validation/NEON/BitwiseXor.cpp
+++ b/tests/validation/NEON/BitwiseXor.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -59,9 +58,9 @@
 Tensor compute_bitwise_xor(const TensorShape &shape)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create xor configure function
     NEBitwiseXor bxor;
@@ -95,9 +94,9 @@
 BOOST_DATA_TEST_CASE(Configuration, SmallShapes() + LargeShapes(), shape)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     BOOST_TEST(src1.info()->is_resizable());
     BOOST_TEST(src2.info()->is_resizable());
diff --git a/tests/validation/NEON/Box3x3.cpp b/tests/validation/NEON/Box3x3.cpp
index c673ff1..0b9c2f0 100644
--- a/tests/validation/NEON/Box3x3.cpp
+++ b/tests/validation/NEON/Box3x3.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -66,8 +65,8 @@
 Tensor compute_box3x3(const TensorShape &shape, BorderMode border_mode, uint8_t constant_border_value)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create and configure function
     NEBox3x3 box3x3;
@@ -98,8 +97,8 @@
 BOOST_DATA_TEST_CASE(Configuration, (SmallShapes() + LargeShapes()) * BorderModes(), shape, border_mode)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/CMakeLists.txt b/tests/validation/NEON/CMakeLists.txt
index 824e0d8..4e15b8e 100644
--- a/tests/validation/NEON/CMakeLists.txt
+++ b/tests/validation/NEON/CMakeLists.txt
@@ -22,7 +22,6 @@
 cmake_minimum_required (VERSION 3.1)
 
 set(arm_compute_test_validation_NEON_SOURCE_FILES
-    ${CMAKE_SOURCE_DIR}/NEON/Helper.h
     ${CMAKE_SOURCE_DIR}/NEON/NEAccessor.h
     ${CMAKE_CURRENT_SOURCE_DIR}/AbsoluteDifference.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/Accumulate.cpp
diff --git a/tests/validation/NEON/ConvolutionLayer.cpp b/tests/validation/NEON/ConvolutionLayer.cpp
index d81834b..da35442 100644
--- a/tests/validation/NEON/ConvolutionLayer.cpp
+++ b/tests/validation/NEON/ConvolutionLayer.cpp
@@ -21,10 +21,11 @@
  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  */
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TypePrinter.h"
 #include "dataset/ConvolutionLayerDataset.h"
+#include "tests/Globals.h"
+#include "tests/Utils.h"
 #include "validation/Datasets.h"
 #include "validation/Reference.h"
 #include "validation/Validation.h"
@@ -51,10 +52,10 @@
                                  const PadStrideInfo &conv_info, int fixed_point_position)
 {
     // Create tensors
-    Tensor src     = create_tensor(input_shape, dt, 1, fixed_point_position);
-    Tensor weights = create_tensor(weights_shape, dt, 1, fixed_point_position);
-    Tensor bias    = create_tensor(bias_shape, dt, 1, fixed_point_position);
-    Tensor dst     = create_tensor(output_shape, dt, 1, fixed_point_position);
+    Tensor src     = create_tensor<Tensor>(input_shape, dt, 1, fixed_point_position);
+    Tensor weights = create_tensor<Tensor>(weights_shape, dt, 1, fixed_point_position);
+    Tensor bias    = create_tensor<Tensor>(bias_shape, dt, 1, fixed_point_position);
+    Tensor dst     = create_tensor<Tensor>(output_shape, dt, 1, fixed_point_position);
 
     // Create and configure function
     NEConvolutionLayer conv;
@@ -107,10 +108,10 @@
     int fixed_point_position = (dt == DataType::F32) ? 0 : 3;
 
     // Create tensors
-    Tensor src     = create_tensor(conv_set.src_shape, dt, 1, fixed_point_position);
-    Tensor weights = create_tensor(conv_set.weights_shape, dt, 1, fixed_point_position);
-    Tensor bias    = create_tensor(conv_set.bias_shape, dt, 1, fixed_point_position);
-    Tensor dst     = create_tensor(conv_set.dst_shape, dt, 1, fixed_point_position);
+    Tensor src     = create_tensor<Tensor>(conv_set.src_shape, dt, 1, fixed_point_position);
+    Tensor weights = create_tensor<Tensor>(conv_set.weights_shape, dt, 1, fixed_point_position);
+    Tensor bias    = create_tensor<Tensor>(conv_set.bias_shape, dt, 1, fixed_point_position);
+    Tensor dst     = create_tensor<Tensor>(conv_set.dst_shape, dt, 1, fixed_point_position);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(weights.info()->is_resizable());
diff --git a/tests/validation/NEON/ConvolutionLayerDirect.cpp b/tests/validation/NEON/ConvolutionLayerDirect.cpp
index 3ddb010..0f5cee6 100644
--- a/tests/validation/NEON/ConvolutionLayerDirect.cpp
+++ b/tests/validation/NEON/ConvolutionLayerDirect.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -69,10 +68,10 @@
                                  DataType dt, PadStrideInfo conv_info, int fixed_point_position = 0)
 {
     // Create tensors
-    Tensor src     = create_tensor(src_shape, dt, 1, fixed_point_position);
-    Tensor weights = create_tensor(weights_shape, dt, 1, fixed_point_position);
-    Tensor bias    = create_tensor(bias_shape, dt, 1, fixed_point_position);
-    Tensor dst     = create_tensor(dst_shape, dt, 1, fixed_point_position);
+    Tensor src     = create_tensor<Tensor>(src_shape, dt, 1, fixed_point_position);
+    Tensor weights = create_tensor<Tensor>(weights_shape, dt, 1, fixed_point_position);
+    Tensor bias    = create_tensor<Tensor>(bias_shape, dt, 1, fixed_point_position);
+    Tensor dst     = create_tensor<Tensor>(dst_shape, dt, 1, fixed_point_position);
 
     // Create and configure function
     NEDirectConvolutionLayer conv_layer;
diff --git a/tests/validation/NEON/DepthConvert.cpp b/tests/validation/NEON/DepthConvert.cpp
index f6ae1f1..1aed9dd 100644
--- a/tests/validation/NEON/DepthConvert.cpp
+++ b/tests/validation/NEON/DepthConvert.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -64,8 +63,8 @@
 Tensor compute_depth_convert(const TensorShape &shape, DataType dt_in, DataType dt_out, ConvertPolicy policy, uint32_t shift, uint32_t fixed_point_position)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, dt_in, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, dt_out, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, dt_in, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, dt_out, 1, fixed_point_position);
 
     // Create and configure function
     NEDepthConvert depth_convert;
@@ -100,8 +99,8 @@
 void compute_configure_validate(const TensorShape &shape, DataType dt_in, DataType dt_out, ConvertPolicy policy, uint32_t shift, uint32_t fixed_point_position)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, dt_in, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, dt_out, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, dt_in, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, dt_out, 1, fixed_point_position);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/FillBorder.cpp b/tests/validation/NEON/FillBorder.cpp
index 9fbeb99..5471092 100644
--- a/tests/validation/NEON/FillBorder.cpp
+++ b/tests/validation/NEON/FillBorder.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -57,7 +56,7 @@
     BorderSize        border_size{ 5 };
 
     // Create tensors
-    Tensor src = create_tensor(TensorShape{ 10U, 10U, 2U }, DataType::U8);
+    Tensor src = create_tensor<Tensor>(TensorShape{ 10U, 10U, 2U }, DataType::U8);
 
     src.info()->extend_padding(padding);
 
diff --git a/tests/validation/NEON/Fixedpoint/Exp_QS16.cpp b/tests/validation/NEON/Fixedpoint/Exp_QS16.cpp
index e6d7d86..bb8961a 100644
--- a/tests/validation/NEON/Fixedpoint/Exp_QS16.cpp
+++ b/tests/validation/NEON/Fixedpoint/Exp_QS16.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -60,8 +59,8 @@
 Tensor compute_exp_qs16(const TensorShape &shape, int fixed_point_position)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::QS16, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, DataType::QS16, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, DataType::QS16, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::QS16, 1, fixed_point_position);
 
     constexpr unsigned int num_elems_processed_per_iteration = 8;
     Window                 window                            = calculate_max_window(*src.info(), Steps(num_elems_processed_per_iteration));
diff --git a/tests/validation/NEON/Fixedpoint/Exp_QS8.cpp b/tests/validation/NEON/Fixedpoint/Exp_QS8.cpp
index f8fc0c2..e901f60 100644
--- a/tests/validation/NEON/Fixedpoint/Exp_QS8.cpp
+++ b/tests/validation/NEON/Fixedpoint/Exp_QS8.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -60,8 +59,8 @@
 Tensor compute_exp_qs8(const TensorShape &shape, int fixed_point_position)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::QS8, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, DataType::QS8, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, DataType::QS8, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::QS8, 1, fixed_point_position);
 
     constexpr unsigned int num_elems_processed_per_iteration = 16;
     Window                 window                            = calculate_max_window(*src.info(), Steps(num_elems_processed_per_iteration));
diff --git a/tests/validation/NEON/Fixedpoint/Invsqrt_QS16.cpp b/tests/validation/NEON/Fixedpoint/Invsqrt_QS16.cpp
index 9211ccf..6b67486 100644
--- a/tests/validation/NEON/Fixedpoint/Invsqrt_QS16.cpp
+++ b/tests/validation/NEON/Fixedpoint/Invsqrt_QS16.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -60,8 +59,8 @@
 Tensor compute_invsqrt_qs16(const TensorShape &shape, int fixed_point_position)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::QS16, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, DataType::QS16, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, DataType::QS16, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::QS16, 1, fixed_point_position);
 
     constexpr unsigned int num_elems_processed_per_iteration = 8;
     Window                 window                            = calculate_max_window(*src.info(), Steps(num_elems_processed_per_iteration));
diff --git a/tests/validation/NEON/Fixedpoint/Invsqrt_QS8.cpp b/tests/validation/NEON/Fixedpoint/Invsqrt_QS8.cpp
index ab63cbe..78719d1 100644
--- a/tests/validation/NEON/Fixedpoint/Invsqrt_QS8.cpp
+++ b/tests/validation/NEON/Fixedpoint/Invsqrt_QS8.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -60,8 +59,8 @@
 Tensor compute_invsqrt_qs8(const TensorShape &shape, int fixed_point_position)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::QS8, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, DataType::QS8, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, DataType::QS8, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::QS8, 1, fixed_point_position);
 
     constexpr unsigned int num_elems_processed_per_iteration = 16;
     Window                 window                            = calculate_max_window(*src.info(), Steps(num_elems_processed_per_iteration));
diff --git a/tests/validation/NEON/Fixedpoint/Log_QS16.cpp b/tests/validation/NEON/Fixedpoint/Log_QS16.cpp
index c23d127..f7a9bc8 100644
--- a/tests/validation/NEON/Fixedpoint/Log_QS16.cpp
+++ b/tests/validation/NEON/Fixedpoint/Log_QS16.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -60,8 +59,8 @@
 Tensor compute_log_qs16(const TensorShape &shape, int fixed_point_position)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::QS16, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, DataType::QS16, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, DataType::QS16, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::QS16, 1, fixed_point_position);
 
     constexpr unsigned int num_elems_processed_per_iteration = 8;
     Window                 window                            = calculate_max_window(*src.info(), Steps(num_elems_processed_per_iteration));
diff --git a/tests/validation/NEON/Fixedpoint/Log_QS8.cpp b/tests/validation/NEON/Fixedpoint/Log_QS8.cpp
index 6789ec7..8cdbd30 100644
--- a/tests/validation/NEON/Fixedpoint/Log_QS8.cpp
+++ b/tests/validation/NEON/Fixedpoint/Log_QS8.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -60,8 +59,8 @@
 Tensor compute_log_qs8(const TensorShape &shape, int fixed_point_position)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::QS8, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, DataType::QS8, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, DataType::QS8, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::QS8, 1, fixed_point_position);
 
     constexpr unsigned int num_elems_processed_per_iteration = 16;
     Window                 window                            = calculate_max_window(*src.info(), Steps(num_elems_processed_per_iteration));
diff --git a/tests/validation/NEON/Fixedpoint/Reciprocal_QS16.cpp b/tests/validation/NEON/Fixedpoint/Reciprocal_QS16.cpp
index c66cf0e..1ef73e2 100644
--- a/tests/validation/NEON/Fixedpoint/Reciprocal_QS16.cpp
+++ b/tests/validation/NEON/Fixedpoint/Reciprocal_QS16.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -60,8 +59,8 @@
 Tensor compute_reciprocal_qs16(const TensorShape &shape, int fixed_point_position)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::QS16, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, DataType::QS16, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, DataType::QS16, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::QS16, 1, fixed_point_position);
 
     constexpr unsigned int num_elems_processed_per_iteration = 8;
     Window                 window                            = calculate_max_window(*src.info(), Steps(num_elems_processed_per_iteration));
diff --git a/tests/validation/NEON/Fixedpoint/Reciprocal_QS8.cpp b/tests/validation/NEON/Fixedpoint/Reciprocal_QS8.cpp
index f1f130a..9c33a9e 100644
--- a/tests/validation/NEON/Fixedpoint/Reciprocal_QS8.cpp
+++ b/tests/validation/NEON/Fixedpoint/Reciprocal_QS8.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -60,8 +59,8 @@
 Tensor compute_reciprocal_qs8(const TensorShape &shape, int fixed_point_position)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::QS8, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, DataType::QS8, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, DataType::QS8, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::QS8, 1, fixed_point_position);
 
     constexpr unsigned int num_elems_processed_per_iteration = 16;
     Window                 window                            = calculate_max_window(*src.info(), Steps(num_elems_processed_per_iteration));
diff --git a/tests/validation/NEON/FullyConnectedLayer.cpp b/tests/validation/NEON/FullyConnectedLayer.cpp
index bda235b..69ff663 100644
--- a/tests/validation/NEON/FullyConnectedLayer.cpp
+++ b/tests/validation/NEON/FullyConnectedLayer.cpp
@@ -21,10 +21,11 @@
  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  */
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TypePrinter.h"
 #include "dataset/FullyConnectedLayerDataset.h"
+#include "tests/Globals.h"
+#include "tests/Utils.h"
 #include "validation/Datasets.h"
 #include "validation/Reference.h"
 #include "validation/Validation.h"
@@ -49,9 +50,9 @@
                                      bool transpose_weights, int fixed_point_position)
 {
     // Create tensors
-    Tensor src  = create_tensor(input_shape, dt, 1, fixed_point_position);
-    Tensor bias = create_tensor(bias_shape, dt, 1, fixed_point_position);
-    Tensor dst  = create_tensor(output_shape, dt, 1, fixed_point_position);
+    Tensor src  = create_tensor<Tensor>(input_shape, dt, 1, fixed_point_position);
+    Tensor bias = create_tensor<Tensor>(bias_shape, dt, 1, fixed_point_position);
+    Tensor dst  = create_tensor<Tensor>(output_shape, dt, 1, fixed_point_position);
 
     // Swap the first and second dimension of weights' shape if transpose_weights is true
     TensorShape ws = weights_shape;
@@ -62,7 +63,7 @@
         ws.set(1, dimx);
     }
 
-    Tensor weights = create_tensor(ws, dt, 1, fixed_point_position);
+    Tensor weights = create_tensor<Tensor>(ws, dt, 1, fixed_point_position);
 
     // Create and configure function.
     // Note: We pass the weights already transposed
@@ -115,9 +116,9 @@
     int fixed_point_position = (dt == DataType::F32) ? 0 : 3;
 
     // Create tensors
-    Tensor src  = create_tensor(fc_set.src_shape, dt, 1, fixed_point_position);
-    Tensor bias = create_tensor(fc_set.bias_shape, dt, 1, fixed_point_position);
-    Tensor dst  = create_tensor(fc_set.dst_shape, dt, 1, fixed_point_position);
+    Tensor src  = create_tensor<Tensor>(fc_set.src_shape, dt, 1, fixed_point_position);
+    Tensor bias = create_tensor<Tensor>(fc_set.bias_shape, dt, 1, fixed_point_position);
+    Tensor dst  = create_tensor<Tensor>(fc_set.dst_shape, dt, 1, fixed_point_position);
 
     // Swap the first and second dimension of weights' shape if transpose_weights is true
     TensorShape ws = fc_set.weights_shape;
@@ -128,7 +129,7 @@
         ws.set(1, dimx);
     }
 
-    Tensor weights = create_tensor(ws, dt, 1, fixed_point_position);
+    Tensor weights = create_tensor<Tensor>(ws, dt, 1, fixed_point_position);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(weights.info()->is_resizable());
diff --git a/tests/validation/NEON/GEMM.cpp b/tests/validation/NEON/GEMM.cpp
index 0172dde..35f65c8 100644
--- a/tests/validation/NEON/GEMM.cpp
+++ b/tests/validation/NEON/GEMM.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -57,10 +56,10 @@
                     const TensorShape &out_shape, float alpha, float beta, DataType dt, int fixed_point_position = 0)
 {
     // Create tensors
-    Tensor src1 = create_tensor(src_shape1, dt, 1, fixed_point_position);
-    Tensor src2 = create_tensor(src_shape2, dt, 1, fixed_point_position);
-    Tensor src3 = create_tensor(src_shape3, dt, 1, fixed_point_position);
-    Tensor dst  = create_tensor(out_shape, dt, 1, fixed_point_position);
+    Tensor src1 = create_tensor<Tensor>(src_shape1, dt, 1, fixed_point_position);
+    Tensor src2 = create_tensor<Tensor>(src_shape2, dt, 1, fixed_point_position);
+    Tensor src3 = create_tensor<Tensor>(src_shape3, dt, 1, fixed_point_position);
+    Tensor dst  = create_tensor<Tensor>(out_shape, dt, 1, fixed_point_position);
 
     // Create and configure function
     NEGEMM gemm;
@@ -112,10 +111,10 @@
     int fixed_point_position = (dt == DataType::F32) ? 0 : 3;
 
     // Create tensors
-    Tensor src1 = create_tensor(gemm_set.shape_a, dt, 1, fixed_point_position);
-    Tensor src2 = create_tensor(gemm_set.shape_b, dt, 1, fixed_point_position);
-    Tensor src3 = create_tensor(gemm_set.shape_c, dt, 1, fixed_point_position);
-    Tensor dst  = create_tensor(gemm_set.shape_d, dt, 1, fixed_point_position);
+    Tensor src1 = create_tensor<Tensor>(gemm_set.shape_a, dt, 1, fixed_point_position);
+    Tensor src2 = create_tensor<Tensor>(gemm_set.shape_b, dt, 1, fixed_point_position);
+    Tensor src3 = create_tensor<Tensor>(gemm_set.shape_c, dt, 1, fixed_point_position);
+    Tensor dst  = create_tensor<Tensor>(gemm_set.shape_d, dt, 1, fixed_point_position);
 
     BOOST_TEST(src1.info()->is_resizable());
     BOOST_TEST(src2.info()->is_resizable());
diff --git a/tests/validation/NEON/Gaussian3x3.cpp b/tests/validation/NEON/Gaussian3x3.cpp
index b69df9a..40ccfe0 100644
--- a/tests/validation/NEON/Gaussian3x3.cpp
+++ b/tests/validation/NEON/Gaussian3x3.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -66,8 +65,8 @@
 Tensor compute_gaussian3x3(const TensorShape &shape, BorderMode border_mode, uint8_t constant_border_value)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create and configure function
     NEGaussian3x3 gaussian3x3;
@@ -98,8 +97,8 @@
 BOOST_DATA_TEST_CASE(Configuration, (SmallShapes() + LargeShapes()) * BorderModes(), shape, border_mode)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/Gaussian5x5.cpp b/tests/validation/NEON/Gaussian5x5.cpp
index bb48961..f8771fb 100644
--- a/tests/validation/NEON/Gaussian5x5.cpp
+++ b/tests/validation/NEON/Gaussian5x5.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -68,8 +67,8 @@
 Tensor compute_gaussian5x5(const TensorShape &shape, BorderMode border_mode, uint8_t constant_border_value)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create and configure function
     NEGaussian5x5 gaussian5x5;
@@ -100,8 +99,8 @@
 BOOST_DATA_TEST_CASE(Configuration, (SmallShapes() + LargeShapes()) * BorderModes(), shape, border_mode)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/IntegralImage.cpp b/tests/validation/NEON/IntegralImage.cpp
index 8789e46..e20eba6 100644
--- a/tests/validation/NEON/IntegralImage.cpp
+++ b/tests/validation/NEON/IntegralImage.cpp
@@ -23,7 +23,6 @@
  */
 
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -60,8 +59,8 @@
 Tensor compute_integral_image(const TensorShape &shape)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U32);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U32);
 
     // Create integral image configure function
     NEIntegralImage integral_image;
@@ -92,8 +91,8 @@
 BOOST_DATA_TEST_CASE(Configuration, SmallShapes() + LargeShapes(), shape)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U32);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U32);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/MeanStdDev.cpp b/tests/validation/NEON/MeanStdDev.cpp
index 7087749..91de5b6 100644
--- a/tests/validation/NEON/MeanStdDev.cpp
+++ b/tests/validation/NEON/MeanStdDev.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -59,7 +58,7 @@
 std::pair<float, float> compute_mean_and_standard_deviation(const TensorShape &shape)
 {
     // Create tensor
-    Tensor src = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create output variables
     float mean    = 0.f;
@@ -92,7 +91,7 @@
 BOOST_DATA_TEST_CASE(Configuration, Small2DShapes() + Large2DShapes(), shape)
 {
     // Create tensor
-    Tensor src = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create output variables
     float mean    = 0.f;
diff --git a/tests/validation/NEON/NonLinearFilter.cpp b/tests/validation/NEON/NonLinearFilter.cpp
index c5d5087..5da7812 100644
--- a/tests/validation/NEON/NonLinearFilter.cpp
+++ b/tests/validation/NEON/NonLinearFilter.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -69,8 +68,8 @@
                                  uint8_t constant_border_value)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create and configure function
     NENonLinearFilter filter;
@@ -113,8 +112,8 @@
     const auto half_mask_size = static_cast<int>(mask_size / 2);
 
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/NormalizationLayer.cpp b/tests/validation/NEON/NormalizationLayer.cpp
index ff791ef..65adaf5 100644
--- a/tests/validation/NEON/NormalizationLayer.cpp
+++ b/tests/validation/NEON/NormalizationLayer.cpp
@@ -21,9 +21,10 @@
  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  */
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TypePrinter.h"
+#include "tests/Globals.h"
+#include "tests/Utils.h"
 #include "validation/Datasets.h"
 #include "validation/Reference.h"
 #include "validation/Validation.h"
@@ -70,8 +71,8 @@
 Tensor compute_normalization_layer(const TensorShape &shape, DataType dt, NormalizationLayerInfo norm_info, int fixed_point_position = 0)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, dt, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, dt, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, dt, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, dt, 1, fixed_point_position);
 
     // Create and configure function
     NENormalizationLayer norm;
diff --git a/tests/validation/NEON/PixelWiseMultiplication.cpp b/tests/validation/NEON/PixelWiseMultiplication.cpp
index fd61c72..8352517 100644
--- a/tests/validation/NEON/PixelWiseMultiplication.cpp
+++ b/tests/validation/NEON/PixelWiseMultiplication.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -67,9 +66,9 @@
                                          int fixed_point_position = 0)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, dt_in0, 1, fixed_point_position);
-    Tensor src2 = create_tensor(shape, dt_in1, 1, fixed_point_position);
-    Tensor dst  = create_tensor(shape, dt_out, 1, fixed_point_position);
+    Tensor src1 = create_tensor<Tensor>(shape, dt_in0, 1, fixed_point_position);
+    Tensor src2 = create_tensor<Tensor>(shape, dt_in1, 1, fixed_point_position);
+    Tensor dst  = create_tensor<Tensor>(shape, dt_out, 1, fixed_point_position);
 
     // Create and configure function
     NEPixelWiseMultiplication multiply;
@@ -131,9 +130,9 @@
                      shape, scale, convert_policy, rounding_policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     validate_configuration(src1, src2, dst, shape, scale, convert_policy, rounding_policy);
 }
@@ -180,9 +179,9 @@
                      shape, scale, convert_policy, rounding_policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor src2 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     validate_configuration(src1, src2, dst, shape, scale, convert_policy, rounding_policy);
 }
@@ -230,9 +229,9 @@
                      shape, dt, scale, convert_policy, rounding_policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, dt);
-    Tensor src2 = create_tensor(shape, DataType::S16);
-    Tensor dst  = create_tensor(shape, DataType::S16);
+    Tensor src1 = create_tensor<Tensor>(shape, dt);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::S16);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::S16);
 
     validate_configuration(src1, src2, dst, shape, scale, convert_policy, rounding_policy);
 }
@@ -277,9 +276,9 @@
                      shape, dt, scale, convert_policy, rounding_policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, dt);
-    Tensor src2 = create_tensor(shape, DataType::S16);
-    Tensor dst  = create_tensor(shape, DataType::S16);
+    Tensor src1 = create_tensor<Tensor>(shape, dt);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::S16);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::S16);
 
     validate_configuration(src1, src2, dst, shape, scale, convert_policy, rounding_policy);
 }
@@ -323,9 +322,9 @@
                      shape, scale, convert_policy, rounding_policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::F32);
-    Tensor src2 = create_tensor(shape, DataType::F32);
-    Tensor dst  = create_tensor(shape, DataType::F32);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::F32);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::F32);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::F32);
 
     validate_configuration(src1, src2, dst, shape, scale, convert_policy, rounding_policy);
 }
@@ -370,9 +369,9 @@
                      shape, scale, convert_policy, rounding_policy)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::F32);
-    Tensor src2 = create_tensor(shape, DataType::F32);
-    Tensor dst  = create_tensor(shape, DataType::F32);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::F32);
+    Tensor src2 = create_tensor<Tensor>(shape, DataType::F32);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::F32);
 
     validate_configuration(src1, src2, dst, shape, scale, convert_policy, rounding_policy);
 }
diff --git a/tests/validation/NEON/PoolingLayer.cpp b/tests/validation/NEON/PoolingLayer.cpp
index b29cd3e..9991437 100644
--- a/tests/validation/NEON/PoolingLayer.cpp
+++ b/tests/validation/NEON/PoolingLayer.cpp
@@ -21,16 +21,16 @@
  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  */
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TypePrinter.h"
 #include "arm_compute/runtime/NEON/functions/NEPoolingLayer.h"
+#include "tests/Globals.h"
+#include "tests/Utils.h"
 #include "tests/dataset/PoolingLayerDataset.h"
 #include "validation/Datasets.h"
 #include "validation/Reference.h"
 #include "validation/Validation.h"
 
-#include <iostream>
 #include <random>
 
 using namespace arm_compute;
@@ -54,8 +54,8 @@
 Tensor compute_pooling_layer(const TensorShape &shape_in, const TensorShape &shape_out, DataType dt, PoolingLayerInfo pool_info, int fixed_point_position = 0)
 {
     // Create tensors
-    Tensor src = create_tensor(shape_in, dt, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape_out, dt, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape_in, dt, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape_out, dt, 1, fixed_point_position);
 
     // Create and configure function
     NEPoolingLayer pool;
diff --git a/tests/validation/NEON/ROIPoolingLayer.cpp b/tests/validation/NEON/ROIPoolingLayer.cpp
index a1d4803..08b051e 100644
--- a/tests/validation/NEON/ROIPoolingLayer.cpp
+++ b/tests/validation/NEON/ROIPoolingLayer.cpp
@@ -25,6 +25,8 @@
 #include "NEON/NEAccessor.h"
 #include "TypePrinter.h"
 #include "arm_compute/runtime/NEON/functions/NEROIPoolingLayer.h"
+#include "tests/Globals.h"
+#include "tests/Utils.h"
 #include "validation/Datasets.h"
 #include "validation/Helpers.h"
 #include "validation/Reference.h"
@@ -50,8 +52,8 @@
     shape_dst.set(3, rois.size());
 
     // Create tensors
-    Tensor     src        = create_tensor(shape, dt);
-    Tensor     dst        = create_tensor(shape_dst, dt);
+    Tensor     src        = create_tensor<Tensor>(shape, dt);
+    Tensor     dst        = create_tensor<Tensor>(shape_dst, dt);
     Array<ROI> rois_array = create_array(rois);
 
     // Create and configure function
diff --git a/tests/validation/NEON/Sobel3x3.cpp b/tests/validation/NEON/Sobel3x3.cpp
index c6c58f1..3b69b2b 100644
--- a/tests/validation/NEON/Sobel3x3.cpp
+++ b/tests/validation/NEON/Sobel3x3.cpp
@@ -23,7 +23,6 @@
  */
 
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "TensorLibrary.h"
 #include "TypePrinter.h"
@@ -66,9 +65,9 @@
 std::pair<Tensor, Tensor> compute_sobel_3x3(const TensorShape &shape, BorderMode border_mode, uint8_t constant_border_value)
 {
     // Create tensors
-    Tensor src   = create_tensor(shape, DataType::U8);
-    Tensor dst_x = create_tensor(shape, DataType::S16);
-    Tensor dst_y = create_tensor(shape, DataType::S16);
+    Tensor src   = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst_x = create_tensor<Tensor>(shape, DataType::S16);
+    Tensor dst_y = create_tensor<Tensor>(shape, DataType::S16);
 
     src.info()->set_format(Format::U8);
     dst_x.info()->set_format(Format::S16);
@@ -105,9 +104,9 @@
 BOOST_DATA_TEST_CASE(Configuration, (SmallShapes() + LargeShapes()) * BorderModes(), shape, border_mode)
 {
     // Create tensors
-    Tensor src   = create_tensor(shape, DataType::U8);
-    Tensor dst_x = create_tensor(shape, DataType::S16);
-    Tensor dst_y = create_tensor(shape, DataType::S16);
+    Tensor src   = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst_x = create_tensor<Tensor>(shape, DataType::S16);
+    Tensor dst_y = create_tensor<Tensor>(shape, DataType::S16);
 
     src.info()->set_format(Format::U8);
     dst_x.info()->set_format(Format::S16);
diff --git a/tests/validation/NEON/Sobel5x5.cpp b/tests/validation/NEON/Sobel5x5.cpp
index 195afa6..456e675 100644
--- a/tests/validation/NEON/Sobel5x5.cpp
+++ b/tests/validation/NEON/Sobel5x5.cpp
@@ -23,7 +23,6 @@
  */
 
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -66,9 +65,9 @@
 std::pair<Tensor, Tensor> compute_sobel_5x5(const TensorShape &shape, BorderMode border_mode, uint8_t constant_border_value)
 {
     // Create tensors
-    Tensor src   = create_tensor(shape, DataType::U8);
-    Tensor dst_x = create_tensor(shape, DataType::S16);
-    Tensor dst_y = create_tensor(shape, DataType::S16);
+    Tensor src   = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst_x = create_tensor<Tensor>(shape, DataType::S16);
+    Tensor dst_y = create_tensor<Tensor>(shape, DataType::S16);
 
     src.info()->set_format(Format::U8);
     dst_x.info()->set_format(Format::S16);
@@ -105,9 +104,9 @@
 BOOST_DATA_TEST_CASE(Configuration, (SmallShapes() + LargeShapes()) * BorderModes(), shape, border_mode)
 {
     // Create tensors
-    Tensor src   = create_tensor(shape, DataType::U8);
-    Tensor dst_x = create_tensor(shape, DataType::S16);
-    Tensor dst_y = create_tensor(shape, DataType::S16);
+    Tensor src   = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst_x = create_tensor<Tensor>(shape, DataType::S16);
+    Tensor dst_y = create_tensor<Tensor>(shape, DataType::S16);
 
     src.info()->set_format(Format::U8);
     dst_x.info()->set_format(Format::S16);
diff --git a/tests/validation/NEON/SoftmaxLayer.cpp b/tests/validation/NEON/SoftmaxLayer.cpp
index 66a9e03..389d448 100644
--- a/tests/validation/NEON/SoftmaxLayer.cpp
+++ b/tests/validation/NEON/SoftmaxLayer.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -66,8 +65,8 @@
 Tensor compute_softmax_layer(const TensorShape &shape, DataType dt, int fixed_point_position = 0)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, dt, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, dt, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, dt, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, dt, 1, fixed_point_position);
 
     // Create and configure function
     NESoftmaxLayer smx_layer;
@@ -111,8 +110,8 @@
     int fixed_point_position = (arm_compute::is_data_type_fixed_point(dt)) ? 3 : 0;
 
     // Create tensors
-    Tensor src = create_tensor(shape, dt, 1, fixed_point_position);
-    Tensor dst = create_tensor(shape, dt, 1, fixed_point_position);
+    Tensor src = create_tensor<Tensor>(shape, dt, 1, fixed_point_position);
+    Tensor dst = create_tensor<Tensor>(shape, dt, 1, fixed_point_position);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/NEON/Threshold.cpp b/tests/validation/NEON/Threshold.cpp
index 55f7889..923245e 100644
--- a/tests/validation/NEON/Threshold.cpp
+++ b/tests/validation/NEON/Threshold.cpp
@@ -22,7 +22,6 @@
  * SOFTWARE.
  */
 #include "Globals.h"
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "PaddingCalculator.h"
 #include "TensorLibrary.h"
@@ -65,8 +64,8 @@
 Tensor compute_threshold(const TensorShape &shape, uint8_t threshold, uint8_t false_value, uint8_t true_value, ThresholdType type, uint8_t upper)
 {
     // Create tensors
-    Tensor src1 = create_tensor(shape, DataType::U8);
-    Tensor dst  = create_tensor(shape, DataType::U8);
+    Tensor src1 = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst  = create_tensor<Tensor>(shape, DataType::U8);
 
     // Create and configure function
     NEThreshold thrsh;
@@ -99,8 +98,8 @@
                      shape, thrshConf)
 {
     // Create tensors
-    Tensor src = create_tensor(shape, DataType::U8);
-    Tensor dst = create_tensor(shape, DataType::U8);
+    Tensor src = create_tensor<Tensor>(shape, DataType::U8);
+    Tensor dst = create_tensor<Tensor>(shape, DataType::U8);
 
     BOOST_TEST(src.info()->is_resizable());
     BOOST_TEST(dst.info()->is_resizable());
diff --git a/tests/validation/system_tests/CL/AlexNet.cpp b/tests/validation/system_tests/CL/AlexNet.cpp
index f7a8820..bbbe4cd 100644
--- a/tests/validation/system_tests/CL/AlexNet.cpp
+++ b/tests/validation/system_tests/CL/AlexNet.cpp
@@ -23,7 +23,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "validation/Validation.h"
 
 #include "arm_compute/runtime/CL/CLSubTensor.h"
diff --git a/tests/validation/system_tests/CL/LeNet5.cpp b/tests/validation/system_tests/CL/LeNet5.cpp
index 8b83cfa..c05690d 100644
--- a/tests/validation/system_tests/CL/LeNet5.cpp
+++ b/tests/validation/system_tests/CL/LeNet5.cpp
@@ -23,7 +23,6 @@
  * SOFTWARE.
  */
 #include "CL/CLAccessor.h"
-#include "CL/Helper.h"
 #include "validation/Validation.h"
 
 #include "arm_compute/runtime/CL/functions/CLActivationLayer.h"
diff --git a/tests/validation/system_tests/NEON/AlexNet.cpp b/tests/validation/system_tests/NEON/AlexNet.cpp
index e56110d..1bb17d4 100644
--- a/tests/validation/system_tests/NEON/AlexNet.cpp
+++ b/tests/validation/system_tests/NEON/AlexNet.cpp
@@ -22,7 +22,6 @@
  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  */
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "validation/Validation.h"
 
diff --git a/tests/validation/system_tests/NEON/LeNet5.cpp b/tests/validation/system_tests/NEON/LeNet5.cpp
index a82b84a..5a311ef 100644
--- a/tests/validation/system_tests/NEON/LeNet5.cpp
+++ b/tests/validation/system_tests/NEON/LeNet5.cpp
@@ -22,7 +22,6 @@
  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  */
-#include "NEON/Helper.h"
 #include "NEON/NEAccessor.h"
 #include "validation/Validation.h"