diff --git a/arm_compute/runtime/CL/functions/CLConvolutionLayer.h b/arm_compute/runtime/CL/functions/CLConvolutionLayer.h
index 2057b6f..cd1ea70 100644
--- a/arm_compute/runtime/CL/functions/CLConvolutionLayer.h
+++ b/arm_compute/runtime/CL/functions/CLConvolutionLayer.h
@@ -34,7 +34,11 @@
 #include "arm_compute/core/CL/kernels/CLIm2ColKernel.h"
 #include "arm_compute/core/CL/kernels/CLWeightsReshapeKernel.h"
 #include "arm_compute/core/Types.h"
+#include "arm_compute/runtime/CL/CLMemoryGroup.h"
 #include "arm_compute/runtime/CL/CLTensor.h"
+#include "arm_compute/runtime/IMemoryManager.h"
+
+#include <memory>
 
 namespace arm_compute
 {
@@ -48,7 +52,7 @@
 {
 public:
     /** Constructor */
-    CLConvolutionLayerReshapeWeights();
+    CLConvolutionLayerReshapeWeights(std::shared_ptr<IMemoryManager> memory_manager = nullptr);
     /** Set the input and output tensors.
      *
      * @param[in]  weights      Weights tensor. Weights are 4D tensor with dimensions [kernel_x, kernel_y, IFM, OFM]. Data type supported: QS8/QS16/F16/F32.
@@ -62,6 +66,7 @@
     void run() override;
 
 private:
+    CLMemoryGroup            _memory_group;
     CLWeightsReshapeKernel   _weights_reshape_kernel;
     CLGEMMTranspose1xWKernel _weights_transposed_kernel;
     CLTensor                 _weights_reshaped;
@@ -81,7 +86,7 @@
 {
 public:
     /** Default constructor */
-    CLConvolutionLayer();
+    CLConvolutionLayer(std::shared_ptr<IMemoryManager> memory_manager = nullptr);
     /** Set the input and output tensors.
      *
      * @param[in]  input        Source tensor. 3 lower dimensions represent a single input [width, height, IFM],
@@ -101,6 +106,7 @@
     void run() override;
 
 private:
+    CLMemoryGroup                    _memory_group;
     CLConvolutionLayerReshapeWeights _reshape_weights;
     CLIm2ColKernel                   _input_im2col_kernel;
     CLGEMMInterleave4x4Kernel        _input_interleave_kernel;
diff --git a/arm_compute/runtime/CL/functions/CLFullyConnectedLayer.h b/arm_compute/runtime/CL/functions/CLFullyConnectedLayer.h
index e076f51..f71e2a3 100644
--- a/arm_compute/runtime/CL/functions/CLFullyConnectedLayer.h
+++ b/arm_compute/runtime/CL/functions/CLFullyConnectedLayer.h
@@ -30,6 +30,7 @@
 #include "arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyKernel.h"
 #include "arm_compute/core/CL/kernels/CLIm2ColKernel.h"
 #include "arm_compute/core/CL/kernels/CLTransposeKernel.h"
+#include "arm_compute/runtime/CL/CLMemoryGroup.h"
 #include "arm_compute/runtime/CL/CLTensor.h"
 
 namespace arm_compute
@@ -64,7 +65,7 @@
 {
 public:
     /** Constructor */
-    CLFullyConnectedLayer();
+    CLFullyConnectedLayer(std::shared_ptr<IMemoryManager> memory_manager = nullptr);
     /** Set the input and output tensors.
      *
      * @param[in]  input                Source tensor. Data type supported: QS8/QS16/F16/F32.
@@ -83,6 +84,7 @@
     void configure_fc_fc(const ICLTensor *input, const ICLTensor *weights, ICLTensor *output);
     void configure_conv_fc(const ICLTensor *input, const ICLTensor *weights, ICLTensor *output);
 
+    CLMemoryGroup                       _memory_group;
     CLIm2ColKernel                      _im2col_kernel;
     CLFullyConnectedLayerReshapeWeights _reshape_weights_kernel;
     CLGEMMMatrixMultiplyKernel          _mm_kernel;
diff --git a/arm_compute/runtime/CL/functions/CLSoftmaxLayer.h b/arm_compute/runtime/CL/functions/CLSoftmaxLayer.h
index 18f7a02..70a265c 100644
--- a/arm_compute/runtime/CL/functions/CLSoftmaxLayer.h
+++ b/arm_compute/runtime/CL/functions/CLSoftmaxLayer.h
@@ -25,8 +25,12 @@
 #define __ARM_COMPUTE_CLSOFTMAXLAYER_H__
 
 #include "arm_compute/core/CL/kernels/CLSoftmaxLayerKernel.h"
+#include "arm_compute/runtime/CL/CLMemoryGroup.h"
 #include "arm_compute/runtime/CL/CLTensor.h"
 #include "arm_compute/runtime/IFunction.h"
+#include "arm_compute/runtime/IMemoryManager.h"
+
+#include <memory>
 
 namespace arm_compute
 {
@@ -46,7 +50,7 @@
 {
 public:
     /** Constructor */
-    CLSoftmaxLayer();
+    CLSoftmaxLayer(std::shared_ptr<IMemoryManager> memory_manager = nullptr);
     /** Set the input and output tensors.
      *
      * @param[in]  input  Source tensor. Data types supported: QS8/QS16/F16/F32
@@ -58,6 +62,7 @@
     void run() override;
 
 private:
+    CLMemoryGroup               _memory_group;
     CLLogits1DMaxKernel         _max_kernel;
     CLLogits1DShiftExpSumKernel _shift_exp_sum_kernel;
     CLLogits1DNormKernel        _norm_kernel;