diff --git a/arm_compute/core/GLES_COMPUTE/GCKernels.h b/arm_compute/core/GLES_COMPUTE/GCKernels.h
index 417c98a..28f6763 100644
--- a/arm_compute/core/GLES_COMPUTE/GCKernels.h
+++ b/arm_compute/core/GLES_COMPUTE/GCKernels.h
@@ -40,6 +40,7 @@
 #include "arm_compute/core/GLES_COMPUTE/kernels/GCGEMMTranspose1xWKernel.h"
 #include "arm_compute/core/GLES_COMPUTE/kernels/GCIm2ColKernel.h"
 #include "arm_compute/core/GLES_COMPUTE/kernels/GCNormalizationLayerKernel.h"
+#include "arm_compute/core/GLES_COMPUTE/kernels/GCNormalizePlanarYUVLayerKernel.h"
 #include "arm_compute/core/GLES_COMPUTE/kernels/GCPixelWiseMultiplicationKernel.h"
 #include "arm_compute/core/GLES_COMPUTE/kernels/GCPoolingLayerKernel.h"
 #include "arm_compute/core/GLES_COMPUTE/kernels/GCSoftmaxLayerKernel.h"
diff --git a/arm_compute/core/GLES_COMPUTE/kernels/GCNormalizePlanarYUVLayerKernel.h b/arm_compute/core/GLES_COMPUTE/kernels/GCNormalizePlanarYUVLayerKernel.h
new file mode 100644
index 0000000..ecf5f44
--- /dev/null
+++ b/arm_compute/core/GLES_COMPUTE/kernels/GCNormalizePlanarYUVLayerKernel.h
@@ -0,0 +1,72 @@
+/*
+ * 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__H__
+#define __ARM_COMPUTE_GCNORMALIZEPLANARYUVLAYERKERNEL_H__
+
+#include "arm_compute/core/GLES_COMPUTE/IGCKernel.h"
+
+namespace arm_compute
+{
+class IGCTensor;
+
+/** Interface for the NormalizePlanarYUV layer kernel.
+ */
+class GCNormalizePlanarYUVLayerKernel : public IGCKernel
+{
+public:
+    /** Constructor */
+    GCNormalizePlanarYUVLayerKernel();
+    /** Prevent instances of this class from being copied (As this class contains pointers) */
+    GCNormalizePlanarYUVLayerKernel(const GCNormalizePlanarYUVLayerKernel &) = delete;
+    /** Prevent instances of this class from being copied (As this class contains pointers) */
+    GCNormalizePlanarYUVLayerKernel &operator=(const GCNormalizePlanarYUVLayerKernel &) = delete;
+    /** Default Move Constructor. */
+    GCNormalizePlanarYUVLayerKernel(GCNormalizePlanarYUVLayerKernel &&) = default;
+    /** Default move assignment operator. */
+    GCNormalizePlanarYUVLayerKernel &operator=(GCNormalizePlanarYUVLayerKernel &&) = default;
+    /** Default destructor */
+    ~GCNormalizePlanarYUVLayerKernel() = default;
+
+    /** Set the input and output tensors.
+     *
+     * @param[in]  input  Source tensor. 3 lower dimensions represent a single input with dimensions [width, height, FM].
+     *                    Data types supported: F16.
+     * @param[out] output Destination tensor. Output will have the same number of dimensions as input. Data type supported: same as @p input
+     * @param[in]  mean   Mean values tensor. 1 dimension with size equal to the feature maps [FM]. Data types supported: Same as @p input
+     * @param[in]  sd     Standard deviation values tensor. 1 dimension with size equal to the feature maps [FM].
+     *                     Data types supported: Same as @p input
+     */
+    void configure(const IGCTensor *input, IGCTensor *output, const IGCTensor *mean, const IGCTensor *sd);
+
+    // Inherited methods overridden:
+    void run(const Window &window) override;
+
+private:
+    const IGCTensor *_input;
+    IGCTensor       *_output;
+    const IGCTensor *_mean;
+    const IGCTensor *_sd;
+};
+}
+#endif /*__ARM_COMPUTE_GCNORMALIZEPLANARYUVLAYERKERNEL_H__ */
diff --git a/arm_compute/runtime/GLES_COMPUTE/GCFunctions.h b/arm_compute/runtime/GLES_COMPUTE/GCFunctions.h
index e76d4ef..e059028 100644
--- a/arm_compute/runtime/GLES_COMPUTE/GCFunctions.h
+++ b/arm_compute/runtime/GLES_COMPUTE/GCFunctions.h
@@ -37,6 +37,7 @@
 #include "arm_compute/runtime/GLES_COMPUTE/functions/GCGEMMInterleave4x4.h"
 #include "arm_compute/runtime/GLES_COMPUTE/functions/GCGEMMTranspose1xW.h"
 #include "arm_compute/runtime/GLES_COMPUTE/functions/GCNormalizationLayer.h"
+#include "arm_compute/runtime/GLES_COMPUTE/functions/GCNormalizePlanarYUVLayer.h"
 #include "arm_compute/runtime/GLES_COMPUTE/functions/GCPixelWiseMultiplication.h"
 #include "arm_compute/runtime/GLES_COMPUTE/functions/GCPoolingLayer.h"
 #include "arm_compute/runtime/GLES_COMPUTE/functions/GCSoftmaxLayer.h"
diff --git a/arm_compute/runtime/GLES_COMPUTE/functions/GCNormalizePlanarYUVLayer.h b/arm_compute/runtime/GLES_COMPUTE/functions/GCNormalizePlanarYUVLayer.h
new file mode 100644
index 0000000..2862eeb
--- /dev/null
+++ b/arm_compute/runtime/GLES_COMPUTE/functions/GCNormalizePlanarYUVLayer.h
@@ -0,0 +1,63 @@
+/*
+ * 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_GCNORMALIZEPLANARYUVLAYER_H__
+#define __ARM_COMPUTE_GCNORMALIZEPLANARYUVLAYER_H__
+
+#include "arm_compute/runtime/IFunction.h"
+
+#include "arm_compute/core/GLES_COMPUTE/kernels/GCNormalizePlanarYUVLayerKernel.h"
+#include "arm_compute/core/Types.h"
+
+namespace arm_compute
+{
+class IGCTensor;
+
+/** Basic function to run @ref GCNormalizePlanarYUVLayerKernel
+ *
+ *  @note The function simulates a NormalizePlanarYUV layer.
+ */
+class GCNormalizePlanarYUVLayer : public IFunction
+{
+public:
+    /** Default constructor */
+    GCNormalizePlanarYUVLayer();
+    /** Set the input and output tensors.
+     *
+     * @param[in]  input  Source tensor. 3 lower dimensions represent a single input with dimensions [width, height, FM].
+     *                    Data types supported: F16.
+     * @param[out] output Destination tensor. Output will have the same number of dimensions as input. Data type supported: same as @p input
+     * @param[in]  mean   Mean values tensor. 1 dimension with size equal to the feature maps [FM]. Data types supported: Same as @p input
+     * @param[in]  sd     Standard deviation values tensor. 1 dimension with size equal to the feature maps [FM].
+     *                     Data types supported: Same as @p input
+     */
+    void configure(const IGCTensor *input, IGCTensor *output, const IGCTensor *mean, const IGCTensor *sd);
+
+    // Inherited methods overridden:
+    void run() override;
+
+private:
+    GCNormalizePlanarYUVLayerKernel _norm_kernel; /**< NormalizePlanarYUV layer kernel to run */
+};
+}
+#endif /* __ARM_COMPUTE_GCNORMALIZEPLANARYUVLAYER_H__ */
diff --git a/src/core/GLES_COMPUTE/GCKernelLibrary.cpp b/src/core/GLES_COMPUTE/GCKernelLibrary.cpp
index e8c6ad2..484c8aa 100644
--- a/src/core/GLES_COMPUTE/GCKernelLibrary.cpp
+++ b/src/core/GLES_COMPUTE/GCKernelLibrary.cpp
@@ -220,6 +220,7 @@
     { "batchnormalization_layer", "batchnormalization_layer.cs" },
     { "concatenate_depth", "concatenate.cs" },
     { "dropout", "dropout.cs" },
+    { "normalize_planar_yuv_layer", "normalize_planar_yuv_layer.cs" },
 };
 
 const std::map<std::string, std::string> GCKernelLibrary::_program_source_map =
@@ -289,6 +290,10 @@
         "dropout.cs",
 #include "./cs_shaders/dropout.csembed"
     },
+    {
+        "normalize_planar_yuv_layer.cs",
+#include "./cs_shaders/normalize_planar_yuv_layer.csembed"
+    },
 #endif /* EMBEDDED_KERNELS */
 };
 
diff --git a/src/core/GLES_COMPUTE/cs_shaders/normalize_planar_yuv_layer.cs b/src/core/GLES_COMPUTE/cs_shaders/normalize_planar_yuv_layer.cs
new file mode 100644
index 0000000..18a9af7
--- /dev/null
+++ b/src/core/GLES_COMPUTE/cs_shaders/normalize_planar_yuv_layer.cs
@@ -0,0 +1,99 @@
+/*
+ * 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.
+ */
+
+layout(local_size_x = LOCAL_SIZE_X, local_size_y = LOCAL_SIZE_Y, local_size_z = LOCAL_SIZE_Z) in;
+
+#include "helpers_cs.h"
+
+precision mediump float;
+
+/** Apply normalize_planar_yuv layer.
+ *
+ * @param[in]  src_ptr    Pointer to the first source tensor. Supported data types: F16
+ * @param[in]  src_attrs  The attributes of the source tensor
+ * @param[out] dst_ptr    Pointer to the destination tensor. Supported data types: same as @p src_ptr
+ * @param[in]  dst_attrs  The attributes of the destination tensor
+ * @param[in]  mean_ptr   Pointer to the mean source tensor. Supported data types: same as @p src_ptr
+ * @param[in]  mean_attrs The attributes of the mean tensor
+ * @param[in]  sd_ptr     Standard deviation values tensor,pointer to the sd tensor. Supported data types: same as @p src_ptr
+ * @param[in]  sd_attrs   The attributes of the sd tensor
+ */
+SHADER_PARAMS_DECLARATION
+{
+    Tensor3DAttributes src_attrs;
+    Tensor3DAttributes dst_attrs;
+    VectorAttributes   mean_attrs;
+    VectorAttributes   sd_attrs;
+};
+
+TENSOR_DECLARATION(1, srcBuffer, uvec2, src_ptr, src_shift, 3, readonly);
+TENSOR_DECLARATION(2, dstBuffer, uvec2, dst_ptr, dst_shift, 3, writeonly);
+TENSOR_DECLARATION(3, meanBuffer, uvec2, mean_ptr, mean_shift, 3, readonly);
+TENSOR_DECLARATION(4, sdBuffer, uvec2, sd_ptr, sd_shift, 3, readonly);
+
+void main(void)
+{
+    Tensor3DIterator src_iter  = CONVERT_TO_TENSOR3D_ITERATOR(src_attrs, src_shift);
+    Tensor3DIterator dst_iter  = CONVERT_TO_TENSOR3D_ITERATOR(dst_attrs, dst_shift);
+    VectorIterator   mean_iter = CONVERT_TO_VECTOR_ITERATOR(mean_attrs, mean_shift);
+    VectorIterator   sd_iter   = CONVERT_TO_VECTOR_ITERATOR(sd_attrs, sd_shift);
+
+    vec4 unpacked_s[3];
+    vec4 tmp;
+    vec4 result;
+
+    uint current_slice = gl_GlobalInvocationID.z;
+    unpacked_s[0]      = LOAD_UNPACK4_CURRENT_ITEM_HALF(src_ptr, src_iter);
+    unpacked_s[1]      = LOAD_UNPACK4_HALF(mean_ptr, TENSOR_OFFSET_ADVANCE_IN_BYTES(mean_iter, current_slice * mean_attrs.stride_x));
+    unpacked_s[2]      = LOAD_UNPACK4_HALF(sd_ptr, TENSOR_OFFSET_ADVANCE_IN_BYTES(sd_iter, current_slice * sd_attrs.stride_x));
+
+    if((current_slice % uint(4)) == uint(0))
+    {
+        tmp    = unpacked_s[0] - unpacked_s[1].x;
+        result = tmp / unpacked_s[2].x;
+
+        STORE_PACK4_CURRENT_ITEM_HALF(dst_ptr, dst_iter, result);
+    }
+    else if((current_slice % uint(4)) == uint(1))
+    {
+        tmp    = unpacked_s[0] - unpacked_s[1].y;
+        result = tmp / unpacked_s[2].y;
+
+        STORE_PACK4_CURRENT_ITEM_HALF(dst_ptr, dst_iter, result);
+    }
+    else if((current_slice % uint(4)) == uint(2))
+    {
+        tmp    = unpacked_s[0] - unpacked_s[1].z;
+        result = tmp / unpacked_s[2].z;
+
+        STORE_PACK4_CURRENT_ITEM_HALF(dst_ptr, dst_iter, result);
+    }
+    else
+    {
+        tmp    = unpacked_s[0] - unpacked_s[1].w;
+        result = tmp / unpacked_s[2].w;
+
+        STORE_PACK4_CURRENT_ITEM_HALF(dst_ptr, dst_iter, result);
+    }
+}
diff --git a/src/core/GLES_COMPUTE/kernels/GCNormalizePlanarYUVLayerKernel.cpp b/src/core/GLES_COMPUTE/kernels/GCNormalizePlanarYUVLayerKernel.cpp
new file mode 100644
index 0000000..bc9c7eb
--- /dev/null
+++ b/src/core/GLES_COMPUTE/kernels/GCNormalizePlanarYUVLayerKernel.cpp
@@ -0,0 +1,114 @@
+/*
+ * 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.
+ */
+#include "arm_compute/core/GLES_COMPUTE/kernels/GCNormalizePlanarYUVLayerKernel.h"
+
+#include "arm_compute/core/AccessWindowStatic.h"
+#include "arm_compute/core/GLES_COMPUTE/GCHelpers.h"
+#include "arm_compute/core/GLES_COMPUTE/GCKernelLibrary.h"
+#include "arm_compute/core/GLES_COMPUTE/IGCTensor.h"
+#include "arm_compute/core/Helpers.h"
+#include "arm_compute/core/Utils.h"
+#include "arm_compute/core/Validate.h"
+#include "arm_compute/core/Window.h"
+
+#include "support/ToolchainSupport.h"
+
+using namespace arm_compute;
+
+GCNormalizePlanarYUVLayerKernel::GCNormalizePlanarYUVLayerKernel()
+    : _input(nullptr), _output(nullptr), _mean(nullptr), _sd(nullptr)
+{
+}
+
+void GCNormalizePlanarYUVLayerKernel::configure(const IGCTensor *input, IGCTensor *output, const IGCTensor *mean, const IGCTensor *sd)
+{
+    ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(input, 1, DataType::F16);
+    ARM_COMPUTE_ERROR_ON_NULLPTR(output);
+    ARM_COMPUTE_ERROR_ON_MISMATCHING_DATA_TYPES(input, output, mean, sd);
+    ARM_COMPUTE_ERROR_ON_MISMATCHING_SHAPES(input, output);
+    ARM_COMPUTE_ERROR_ON_MISMATCHING_SHAPES(mean, sd);
+    ARM_COMPUTE_ERROR_ON(input->info()->dimension(2) != mean->info()->dimension(0));
+
+    _input  = input;
+    _output = output;
+    _mean   = mean;
+    _sd     = sd;
+
+    const unsigned int num_elems_processed_per_iteration = 4;
+
+    // Set build options
+    std::set<std::string> build_opts;
+    build_opts.emplace(("#define LOCAL_SIZE_X " + support::cpp11::to_string(1)));
+    build_opts.emplace(("#define LOCAL_SIZE_Y " + support::cpp11::to_string(1)));
+    build_opts.emplace(("#define LOCAL_SIZE_Z " + support::cpp11::to_string(1)));
+
+    // Create kernel
+    _kernel = static_cast<GCKernel>(GCKernelLibrary::get().create_kernel("normalize_planar_yuv_layer", build_opts));
+
+    // Configure kernel window
+    Window win = calculate_max_window(*input->info(), Steps(num_elems_processed_per_iteration));
+
+    AccessWindowHorizontal input_access(input->info(), 0, num_elems_processed_per_iteration);
+    AccessWindowHorizontal output_access(output->info(), 0, num_elems_processed_per_iteration);
+    const int              mean_padding = ceil_to_multiple(mean->info()->dimension(0), num_elems_processed_per_iteration) - mean->info()->dimension(0);
+    const int              sd_padding   = ceil_to_multiple(sd->info()->dimension(0), num_elems_processed_per_iteration) - sd->info()->dimension(0);
+    AccessWindowStatic     mean_access(mean->info(), 0, 0, mean->info()->dimension(0) + mean_padding, mean->info()->dimension(1));
+    AccessWindowStatic     sd_access(sd->info(), 0, 0, sd->info()->dimension(0) + sd_padding, sd->info()->dimension(1));
+
+    update_window_and_padding(win, input_access, output_access, mean_access, sd_access);
+    output_access.set_valid_region(win, input->info()->valid_region());
+
+    IGCKernel::configure(win);
+}
+
+void GCNormalizePlanarYUVLayerKernel::run(const Window &window)
+{
+    ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this);
+    ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(IKernel::window(), window);
+
+    _kernel.use();
+
+    Window slice = window.first_slice_window_3D();
+
+    Window slice_in;
+    //slice_in.use_tensor_dimensions(_mean->info()->tensor_shape());
+    slice_in = window.first_slice_window_1D();
+    slice_in.set(Window::DimX, Window::Dimension(0, 0, 0));
+
+    unsigned int idx = 2 * num_arguments_per_3D_tensor();
+    add_1D_tensor_argument(idx, _mean, 3, slice_in);
+    add_1D_tensor_argument(idx, _sd, 4, slice_in);
+
+    do
+    {
+        idx = 0;
+        add_3D_tensor_argument(idx, _input, 1, slice);
+        add_3D_tensor_argument(idx, _output, 2, slice);
+
+        _kernel.update_shader_params();
+
+        enqueue(*this, slice);
+    }
+    while(window.slide_window_slice_3D(slice));
+}
diff --git a/src/runtime/GLES_COMPUTE/functions/GCNormalizePlanarYUVLayer.cpp b/src/runtime/GLES_COMPUTE/functions/GCNormalizePlanarYUVLayer.cpp
new file mode 100755
index 0000000..de04a28
--- /dev/null
+++ b/src/runtime/GLES_COMPUTE/functions/GCNormalizePlanarYUVLayer.cpp
@@ -0,0 +1,48 @@
+/*
+ * 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.
+ */
+
+#include "arm_compute/runtime/GLES_COMPUTE/functions/GCNormalizePlanarYUVLayer.h"
+
+#include "arm_compute/core/Error.h"
+#include "arm_compute/core/TensorInfo.h"
+#include "arm_compute/core/Types.h"
+#include "arm_compute/core/Validate.h"
+#include "arm_compute/runtime/GLES_COMPUTE/GCScheduler.h"
+
+using namespace arm_compute;
+
+GCNormalizePlanarYUVLayer::GCNormalizePlanarYUVLayer()
+    : _norm_kernel()
+{
+}
+
+void GCNormalizePlanarYUVLayer::configure(const IGCTensor *input, IGCTensor *output, const IGCTensor *mean, const IGCTensor *sd)
+{
+    _norm_kernel.configure(input, output, mean, sd);
+}
+
+void GCNormalizePlanarYUVLayer::run()
+{
+    GCScheduler::get().enqueue(_norm_kernel, true);
+}
diff --git a/tests/benchmark/GLES_COMPUTE/NormalizePlanarYUVLayer.cpp b/tests/benchmark/GLES_COMPUTE/NormalizePlanarYUVLayer.cpp
new file mode 100644
index 0000000..3324180
--- /dev/null
+++ b/tests/benchmark/GLES_COMPUTE/NormalizePlanarYUVLayer.cpp
@@ -0,0 +1,56 @@
+/*
+ * 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,
+ * FITCLSS 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 CONCLCTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "arm_compute/core/TensorShape.h"
+#include "arm_compute/core/Types.h"
+#include "arm_compute/runtime/GLES_COMPUTE/GCTensor.h"
+#include "arm_compute/runtime/GLES_COMPUTE/GCTensorAllocator.h"
+#include "arm_compute/runtime/GLES_COMPUTE/functions/GCNormalizePlanarYUVLayer.h"
+#include "tests/GLES_COMPUTE/GCAccessor.h"
+#include "tests/benchmark/fixtures/NormalizePlanarYUVLayerFixture.h"
+#include "tests/datasets/system_tests/googlenet/inceptionv4/GoogLeNetInceptionV4NormalizePlanarYUVLayerDataset.h"
+#include "tests/framework/Macros.h"
+#include "tests/framework/datasets/Datasets.h"
+#include "utils/TypePrinter.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace
+{
+const auto data_types = framework::dataset::make("DataType", { DataType::F16 });
+} // namespace
+
+using GCNormalizePlanarYUVLayerFixture = NormalizePlanarYUVLayerFixture<GCTensor, GCNormalizePlanarYUVLayer, GCAccessor>;
+
+TEST_SUITE(GC)
+
+REGISTER_FIXTURE_DATA_TEST_CASE(GoogLeNetInceptionV4NormalizePlanarYUVLayer, GCNormalizePlanarYUVLayerFixture, framework::DatasetMode::ALL,
+                                framework::dataset::combine(framework::dataset::combine(datasets::GoogLeNetInceptionV4NormalizePlanarYUVLayerDataset(),
+                                                                                        data_types),
+                                                            framework::dataset::make("Batches", 1)));
+
+TEST_SUITE_END()
+} // namespace test
+} // namespace arm_compute
diff --git a/tests/benchmark/fixtures/NormalizePlanarYUVLayerFixture.h b/tests/benchmark/fixtures/NormalizePlanarYUVLayerFixture.h
new file mode 100644
index 0000000..9234353
--- /dev/null
+++ b/tests/benchmark/fixtures/NormalizePlanarYUVLayerFixture.h
@@ -0,0 +1,103 @@
+/*
+ * 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_NORMALIZEPLANARYUVLAYERFIXTURE
+#define ARM_COMPUTE_TEST_NORMALIZEPLANARYUVLAYERFIXTURE
+
+#include "arm_compute/core/TensorShape.h"
+#include "arm_compute/core/Types.h"
+#include "tests/Globals.h"
+#include "tests/Utils.h"
+#include "tests/framework/Fixture.h"
+#ifdef ARM_COMPUTE_GC
+#include "arm_compute/runtime/GLES_COMPUTE/GCScheduler.h"
+#include "tests/GLES_COMPUTE/Helper.h"
+
+using namespace arm_compute::test::gles_compute;
+#endif /* ARM_COMPUTE_GC */
+
+namespace arm_compute
+{
+namespace test
+{
+/** Fixture that can be used for NEON and CL */
+template <typename TensorType, typename Function, typename Accessor>
+class NormalizePlanarYUVLayerFixture : public framework::Fixture
+{
+public:
+    template <typename...>
+    void setup(TensorShape tensor_shape, TensorShape param_shape, DataType data_type, int batches)
+    {
+        // Set batched in source and destination shapes
+        tensor_shape.set(tensor_shape.num_dimensions(), batches);
+
+        // Create tensors
+        src  = create_tensor<TensorType>(tensor_shape, data_type, 1);
+        dst  = create_tensor<TensorType>(tensor_shape, data_type, 1);
+        mean = create_tensor<TensorType>(param_shape, data_type, 1);
+        sd   = create_tensor<TensorType>(param_shape, data_type, 1);
+
+        // Create and configure function
+        normalize_planar_yuv_layer.configure(&src, &dst, &mean, &sd);
+
+        // Allocate tensors
+        src.allocator()->allocate();
+        dst.allocator()->allocate();
+        mean.allocator()->allocate();
+        sd.allocator()->allocate();
+
+        // Fill tensors
+        library->fill_tensor_uniform(Accessor(src), 0);
+        library->fill_tensor_uniform(Accessor(mean), 1);
+        library->fill_tensor_uniform(Accessor(sd), 2);
+    }
+
+    void run()
+    {
+        normalize_planar_yuv_layer.run();
+#ifdef ARM_COMPUTE_GC
+        if(opengles31_is_available() && std::is_same<typename std::decay<TensorType>::type, arm_compute::GCTensor>::value)
+        {
+            force_sync_tensor(dst);
+        }
+#endif /* ARM_COMPUTE_GC */
+    }
+
+    void teardown()
+    {
+        src.allocator()->free();
+        dst.allocator()->free();
+        mean.allocator()->free();
+        sd.allocator()->free();
+    }
+
+private:
+    TensorType src{};
+    TensorType dst{};
+    TensorType mean{};
+    TensorType sd{};
+    Function   normalize_planar_yuv_layer{};
+};
+} // namespace test
+} // namespace arm_compute
+#endif /* ARM_COMPUTE_TEST_NORMALIZEPLANARYUVLAYERFIXTURE */
diff --git a/tests/datasets/NormalizePlanarYUVLayerDataset.h b/tests/datasets/NormalizePlanarYUVLayerDataset.h
new file mode 100644
index 0000000..2d71a56
--- /dev/null
+++ b/tests/datasets/NormalizePlanarYUVLayerDataset.h
@@ -0,0 +1,107 @@
+/*
+ * 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_NORMALIZE_PLANAR_YUV_LAYER_DATASET
+#define ARM_COMPUTE_TEST_NORMALIZE_PLANAR_YUV_LAYER_DATASET
+
+#include "utils/TypePrinter.h"
+
+#include "arm_compute/core/TensorShape.h"
+#include "arm_compute/core/Types.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace datasets
+{
+class NormalizePlanarYUVLayerDataset
+{
+public:
+    using type = std::tuple<TensorShape, TensorShape>;
+
+    struct iterator
+    {
+        iterator(std::vector<TensorShape>::const_iterator tensor_it,
+                 std::vector<TensorShape>::const_iterator param_it)
+            : _tensor_it{ std::move(tensor_it) },
+              _param_it{ std::move(param_it) }
+        {
+        }
+
+        std::string description() const
+        {
+            std::stringstream description;
+            description << "In=" << *_tensor_it << ":";
+            description << "Out=" << *_tensor_it << ":";
+            description << "Mean=" << *_param_it << ":";
+            description << "Sd=" << *_param_it << ":";
+            return description.str();
+        }
+
+        NormalizePlanarYUVLayerDataset::type operator*() const
+        {
+            return std::make_tuple(*_tensor_it, *_param_it);
+        }
+
+        iterator &operator++()
+        {
+            ++_tensor_it;
+            ++_param_it;
+
+            return *this;
+        }
+
+    private:
+        std::vector<TensorShape>::const_iterator _tensor_it;
+        std::vector<TensorShape>::const_iterator _param_it;
+    };
+
+    iterator begin() const
+    {
+        return iterator(_tensor_shapes.begin(), _param_shapes.begin());
+    }
+
+    int size() const
+    {
+        return std::min(_tensor_shapes.size(), _param_shapes.size());
+    }
+
+    void add_config(TensorShape tensor, TensorShape param)
+    {
+        _tensor_shapes.emplace_back(std::move(tensor));
+        _param_shapes.emplace_back(std::move(param));
+    }
+
+protected:
+    NormalizePlanarYUVLayerDataset()                                  = default;
+    NormalizePlanarYUVLayerDataset(NormalizePlanarYUVLayerDataset &&) = default;
+
+private:
+    std::vector<TensorShape> _tensor_shapes{};
+    std::vector<TensorShape> _param_shapes{};
+};
+} // namespace datasets
+} // namespace test
+} // namespace arm_compute
+#endif /* ARM_COMPUTE_TEST_NORMALIZE_PLANAR_YUV_LAYER_DATASET */
diff --git a/tests/datasets/RandomNormalizePlanarYUVLayerDataset.h b/tests/datasets/RandomNormalizePlanarYUVLayerDataset.h
new file mode 100644
index 0000000..5693004
--- /dev/null
+++ b/tests/datasets/RandomNormalizePlanarYUVLayerDataset.h
@@ -0,0 +1,54 @@
+/*
+ * 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_RANDOM_NORMALIZE_PLANAR_YUV_LAYER_DATASET
+#define ARM_COMPUTE_TEST_RANDOM_NORMALIZE_PLANAR_YUV_LAYER_DATASET
+
+#include "tests/datasets/NormalizePlanarYUVLayerDataset.h"
+
+#include "utils/TypePrinter.h"
+
+#include "arm_compute/core/TensorShape.h"
+#include "arm_compute/core/Types.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace datasets
+{
+class RandomNormalizePlanarYUVLayerDataset final : public NormalizePlanarYUVLayerDataset
+{
+public:
+    RandomNormalizePlanarYUVLayerDataset()
+    {
+        add_config(TensorShape(15U, 4U, 4U, 1U), TensorShape(4U));
+        add_config(TensorShape(21U, 11U, 12U, 1U), TensorShape(12U));
+        add_config(TensorShape(7U, 3U, 6U, 1U), TensorShape(6U));
+        add_config(TensorShape(7U, 2U, 3U, 1U), TensorShape(3U));
+    }
+};
+} // namespace datasets
+} // namespace test
+} // namespace arm_compute
+#endif /* ARM_COMPUTE_TEST_RANDOM_NORMALIZE_PLANAR_YUV_LAYER_DATASET */
diff --git a/tests/datasets/system_tests/googlenet/inceptionv4/GoogLeNetInceptionV4NormalizePlanarYUVLayerDataset.h b/tests/datasets/system_tests/googlenet/inceptionv4/GoogLeNetInceptionV4NormalizePlanarYUVLayerDataset.h
new file mode 100644
index 0000000..e0da484
--- /dev/null
+++ b/tests/datasets/system_tests/googlenet/inceptionv4/GoogLeNetInceptionV4NormalizePlanarYUVLayerDataset.h
@@ -0,0 +1,94 @@
+/*
+ * 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_GOOGLENETINCEPTIONV4_NORMALIZEPLANARYUV_LAYER_DATASET
+#define ARM_COMPUTE_TEST_GOOGLENETINCEPTIONV4_NORMALIZEPLANARYUV_LAYER_DATASET
+
+#include "tests/datasets/NormalizePlanarYUVLayerDataset.h"
+
+#include "utils/TypePrinter.h"
+
+#include "arm_compute/core/TensorShape.h"
+#include "arm_compute/core/Types.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace datasets
+{
+class GoogLeNetInceptionV4NormalizePlanarYUVLayerDataset final : public NormalizePlanarYUVLayerDataset
+{
+public:
+    GoogLeNetInceptionV4NormalizePlanarYUVLayerDataset()
+    {
+        // conv1_3x3_s2_bn
+        add_config(TensorShape(149U, 149U, 32U), TensorShape(32U));
+        // conv2_3x3_s1_bn
+        add_config(TensorShape(147U, 147U, 32U), TensorShape(32U));
+        // conv3_3x3_s1_bn
+        add_config(TensorShape(147U, 147U, 64U), TensorShape(64U));
+        // inception_stem1_3x3_s2_bn
+        add_config(TensorShape(73U, 73U, 96U), TensorShape(96U));
+        // inception_stem2_3x3_reduce_bn, inception_stem2_1x7_reduce_bn, inception_stem2_1x7_bn, inception_stem2_7x1_bn
+        add_config(TensorShape(73U, 73U, 64U), TensorShape(64U));
+        // inception_stem2_3x3_bn, inception_stem2_3x3_2_bn
+        add_config(TensorShape(71U, 71U, 96U), TensorShape(96U));
+        // inception_stem3_3x3_s2_bn, reduction_a_3x3_2_reduce_bn
+        add_config(TensorShape(35U, 35U, 192U), TensorShape(192U));
+        // inception_a1_1x1_2_bn, inception_a1_3x3_bn, inception_a1_3x3_2_bn, inception_a1_3x3_3_bn, inception_a1_1x1_bn, inception_a2_1x1_2_bn, inception_a2_3x3_bn, inception_a2_3x3_2_bn, inception_a2_3x3_3_bn, inception_a2_1x1_bn, inception_a3_1x1_2_bn, inception_a3_3x3_bn, inception_a3_3x3_2_bn, inception_a3_3x3_3_bn, inception_a3_1x1_bn, inception_a4_1x1_2_bn, inception_a4_3x3_bn, inception_a4_3x3_2_bn, inception_a4_3x3_3_bn, inception_a4_1x1_bn
+        add_config(TensorShape(35U, 35U, 96U), TensorShape(96U));
+        // inception_a1_3x3_reduce_bn, inception_a1_3x3_2_reduce_bn, inception_a2_3x3_reduce_bn, inception_a2_3x3_2_reduce_bn, inception_a3_3x3_reduce_bn, inception_a3_3x3_2_reduce_bn, inception_a4_3x3_reduce_bn, inception_a4_3x3_2_reduce_bn
+        add_config(TensorShape(35U, 35U, 64U), TensorShape(64U));
+        // reduction_a_3x3_bn, inception_b1_1x1_2_bn, inception_b2_1x1_2_bn, inception_b3_1x1_2_bn, inception_b4_1x1_2_bn, inception_b5_1x1_2_bn, inception_b6_1x1_2_bn, inception_b7_1x1_2_bn
+        add_config(TensorShape(17U, 17U, 384U), TensorShape(384U));
+        // reduction_a_3x3_2_bn
+        add_config(TensorShape(35U, 35U, 224U), TensorShape(224U));
+        // reduction_a_3x3_3_bn, inception_b1_7x1_bn, inception_b1_1x7_3_bn, inception_b2_7x1_bn, inception_b2_1x7_3_bn, inception_b3_7x1_bn, inception_b3_1x7_3_bn, inception_b4_7x1_bn, inception_b4_1x7_3_bn, inception_b5_7x1_bn, inception_b5_1x7_3_bn, inception_b6_7x1_bn, inception_b6_1x7_3_bn, inception_b7_7x1_bn, inception_b7_1x7_3_bn, reduction_b_1x7_reduce_bn, reduction_b_1x7_bn
+        add_config(TensorShape(17U, 17U, 256U), TensorShape(256U));
+        // inception_b1_1x7_reduce_bn, inception_b1_7x1_2_reduce_bn, inception_b1_7x1_2_bn, inception_b2_1x7_reduce_bn, inception_b2_7x1_2_reduce_bn, inception_b2_7x1_2_bn, inception_b3_1x7_reduce_bn, inception_b3_7x1_2_reduce_bn, inception_b3_7x1_2_bn, inception_b4_1x7_reduce_bn, inception_b4_7x1_2_reduce_bn, inception_b4_7x1_2_bn, inception_b5_1x7_reduce_bn, inception_b5_7x1_2_reduce_bn, inception_b5_7x1_2_bn, inception_b6_1x7_reduce_bn, inception_b6_7x1_2_reduce_bn, inception_b6_7x1_2_bn, inception_b7_1x7_reduce_bn, inception_b7_7x1_2_reduce_bn, inception_b7_7x1_2_bn, reduction_b_3x3_reduce_bn
+        add_config(TensorShape(17U, 17U, 192U), TensorShape(192U));
+        // inception_b1_1x7_bn, inception_b1_1x7_2_bn, inception_b1_7x1_3_bn, inception_b2_1x7_bn, inception_b2_1x7_2_bn, inception_b2_7x1_3_bn, inception_b3_1x7_bn, inception_b3_1x7_2_bn, inception_b3_7x1_3_bn, inception_b4_1x7_bn, inception_b4_1x7_2_bn, inception_b4_7x1_3_bn, inception_b5_1x7_bn, inception_b5_1x7_2_bn, inception_b5_7x1_3_bn, inception_b6_1x7_bn, inception_b6_1x7_2_bn, inception_b6_7x1_3_bn, inception_b7_1x7_bn, inception_b7_1x7_2_bn, inception_b7_7x1_3_bn
+        add_config(TensorShape(17U, 17U, 224U), TensorShape(224U));
+        // inception_b1_1x1_bn, inception_b2_1x1_bn, inception_b3_1x1_bn, inception_b4_1x1_bn, inception_b5_1x1_bn, inception_b6_1x1_bn, inception_b7_1x1_bn
+        add_config(TensorShape(17U, 17U, 128U), TensorShape(128U));
+        // reduction_b_3x3_bn
+        add_config(TensorShape(8U, 8U, 192U), TensorShape(192U));
+        // reduction_b_7x1_bn
+        add_config(TensorShape(17U, 17U, 320U), TensorShape(320U));
+        // reduction_b_3x3_2_bn
+        add_config(TensorShape(8U, 8U, 320U), TensorShape(320U));
+        // inception_c1_1x1_2_bn, inception_c1_1x3_bn, inception_c1_3x1_bn, inception_c1_1x3_3_bn, inception_c1_3x1_3_bn, inception_c1_1x1_bn, inception_c2_1x1_2_bn, inception_c2_1x3_bn, inception_c2_3x1_bn, inception_c2_1x3_3_bn, inception_c2_3x1_3_bn, inception_c2_1x1_bn, inception_c3_1x1_2_bn, inception_c3_1x3_bn, inception_c3_3x1_bn, inception_c3_1x3_3_bn, inception_c3_3x1_3_bn, inception_c3_1x1_bn
+        add_config(TensorShape(8U, 8U, 256U), TensorShape(256U));
+        // inception_c1_1x1_3_bn, inception_c1_1x1_4_bn, inception_c2_1x1_3_bn, inception_c2_1x1_4_bn, inception_c3_1x1_3_bn, inception_c3_1x1_4_bn
+        add_config(TensorShape(8U, 8U, 384U), TensorShape(384U));
+        // inception_c1_3x1_2_bn, inception_c2_3x1_2_bn, inception_c3_3x1_2_bn
+        add_config(TensorShape(8U, 8U, 448U), TensorShape(448U));
+        // inception_c1_1x3_2_bn, inception_c2_1x3_2_bn, inception_c3_1x3_2_bn
+        add_config(TensorShape(8U, 8U, 512U), TensorShape(512U));
+    }
+};
+} // namespace datasets
+} // namespace test
+} // namespace arm_compute
+#endif /* ARM_COMPUTE_TEST_GOOGLENETINCEPTIONV4_NORMALIZEPLANARYUV_LAYER_DATASET */
diff --git a/tests/validation/GLES_COMPUTE/NormalizePlanarYUVLayer.cpp b/tests/validation/GLES_COMPUTE/NormalizePlanarYUVLayer.cpp
new file mode 100644
index 0000000..e06b19c
--- /dev/null
+++ b/tests/validation/GLES_COMPUTE/NormalizePlanarYUVLayer.cpp
@@ -0,0 +1,88 @@
+/*
+ * 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.
+ */
+#include "arm_compute/core/Types.h"
+#include "arm_compute/runtime/GLES_COMPUTE/GCTensor.h"
+#include "arm_compute/runtime/GLES_COMPUTE/GCTensorAllocator.h"
+#include "arm_compute/runtime/GLES_COMPUTE/functions/GCNormalizePlanarYUVLayer.h"
+#include "tests/GLES_COMPUTE/GCAccessor.h"
+#include "tests/PaddingCalculator.h"
+#include "tests/datasets/RandomNormalizePlanarYUVLayerDataset.h"
+#include "tests/datasets/ShapeDatasets.h"
+#include "tests/framework/Asserts.h"
+#include "tests/framework/Macros.h"
+#include "tests/framework/datasets/Datasets.h"
+#include "tests/validation/Validation.h"
+#include "tests/validation/fixtures/NormalizePlanarYUVLayerFixture.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+namespace
+{
+constexpr AbsoluteTolerance<float> tolerance_f16(0.5f); /**< Tolerance value for comparing reference's output against implementation's output for DataType::F16 */
+} // namespace
+
+TEST_SUITE(GC)
+TEST_SUITE(NormalizePlanarYUVLayer)
+
+template <typename T>
+using GCNormalizePlanarYUVLayerFixture = NormalizePlanarYUVLayerValidationFixture<GCTensor, GCAccessor, GCNormalizePlanarYUVLayer, T>;
+
+DATA_TEST_CASE(Configuration, framework::DatasetMode::ALL, combine(datasets::RandomNormalizePlanarYUVLayerDataset(), framework::dataset::make("DataType", { DataType::F16 })),
+               shape0, shape1, dt)
+{
+    // Create tensors
+    GCTensor src  = create_tensor<GCTensor>(shape0, dt, 1);
+    GCTensor dst  = create_tensor<GCTensor>(shape0, dt, 1);
+    GCTensor mean = create_tensor<GCTensor>(shape1, dt, 1);
+    GCTensor sd   = create_tensor<GCTensor>(shape1, dt, 1);
+
+    // Create and Configure function
+    GCNormalizePlanarYUVLayer norm;
+    norm.configure(&src, &dst, &mean, &sd);
+
+    // Validate valid region
+    const ValidRegion valid_region = shape_to_valid_region(shape0);
+    validate(dst.info()->valid_region(), valid_region);
+}
+
+TEST_SUITE(Float)
+TEST_SUITE(FP16)
+FIXTURE_DATA_TEST_CASE(Random, GCNormalizePlanarYUVLayerFixture<half>, framework::DatasetMode::PRECOMMIT, combine(datasets::RandomNormalizePlanarYUVLayerDataset(),
+                                                                                                                  framework::dataset::make("DataType", DataType::F16)))
+{
+    // Validate output
+    validate(GCAccessor(_target), _reference, tolerance_f16, 0);
+}
+TEST_SUITE_END()
+TEST_SUITE_END()
+
+TEST_SUITE_END()
+TEST_SUITE_END()
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
diff --git a/tests/validation/Helpers.h b/tests/validation/Helpers.h
old mode 100644
new mode 100755
index ab7dbd8..ba45968
--- a/tests/validation/Helpers.h
+++ b/tests/validation/Helpers.h
@@ -202,6 +202,20 @@
     return bounds;
 }
 
+/** Helper function to get the testing range for NormalizePlanarYUV layer.
+ *
+ * @return A pair containing the lower upper testing bounds.
+ */
+template <typename T>
+std::pair<T, T> get_normalize_planar_yuv_layer_test_bounds()
+{
+    std::pair<T, T> bounds;
+
+    bounds = std::make_pair(-1.f, 1.f);
+
+    return bounds;
+}
+
 /** Convert quantized simple tensor into float using tensor quantization information.
  *
  * @param[in] src Quantized tensor.
diff --git a/tests/validation/fixtures/NormalizePlanarYUVLayerFixture.h b/tests/validation/fixtures/NormalizePlanarYUVLayerFixture.h
new file mode 100644
index 0000000..ae5c53a
--- /dev/null
+++ b/tests/validation/fixtures/NormalizePlanarYUVLayerFixture.h
@@ -0,0 +1,140 @@
+/*
+ * 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_NORMALIZE_PLANAR_YUV_LAYER_FIXTURE
+#define ARM_COMPUTE_TEST_NORMALIZE_PLANAR_YUV_LAYER_FIXTURE
+
+#include "arm_compute/core/TensorShape.h"
+#include "arm_compute/core/Types.h"
+#include "tests/AssetsLibrary.h"
+#include "tests/Globals.h"
+#include "tests/IAccessor.h"
+#include "tests/framework/Asserts.h"
+#include "tests/framework/Fixture.h"
+#include "tests/validation/Helpers.h"
+#include "tests/validation/reference/NormalizePlanarYUVLayer.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+template <typename TensorType, typename AccessorType, typename FunctionType, typename T>
+class NormalizePlanarYUVLayerValidationFixedPointFixture : public framework::Fixture
+{
+public:
+    template <typename...>
+    void setup(TensorShape shape0, TensorShape shape1, DataType dt)
+    {
+        _data_type = dt;
+        _target    = compute_target(shape0, shape1, dt);
+        _reference = compute_reference(shape0, shape1, dt);
+    }
+
+protected:
+    template <typename U>
+    void fill(U &&src_tensor, U &&mean_tensor, U &&sd_tensor)
+    {
+        if(is_data_type_float(_data_type))
+        {
+            float min_bound = 0.f;
+            float max_bound = 0.f;
+            std::tie(min_bound, max_bound) = get_normalize_planar_yuv_layer_test_bounds<T>();
+            std::uniform_real_distribution<> distribution(min_bound, max_bound);
+            std::uniform_real_distribution<> distribution_sd(0, max_bound);
+            library->fill(src_tensor, distribution, 0);
+            library->fill(mean_tensor, distribution, 1);
+            library->fill(sd_tensor, distribution_sd, 2);
+        }
+    }
+
+    TensorType compute_target(const TensorShape &shape0, const TensorShape &shape1, DataType dt)
+    {
+        // Create tensors
+        TensorType src  = create_tensor<TensorType>(shape0, dt, 1);
+        TensorType dst  = create_tensor<TensorType>(shape0, dt, 1);
+        TensorType mean = create_tensor<TensorType>(shape1, dt, 1);
+        TensorType sd   = create_tensor<TensorType>(shape1, dt, 1);
+
+        // Create and configure function
+        FunctionType norm;
+        norm.configure(&src, &dst, &mean, &sd);
+
+        ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
+        ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
+        ARM_COMPUTE_EXPECT(mean.info()->is_resizable(), framework::LogLevel::ERRORS);
+        ARM_COMPUTE_EXPECT(sd.info()->is_resizable(), framework::LogLevel::ERRORS);
+
+        // Allocate tensors
+        src.allocator()->allocate();
+        dst.allocator()->allocate();
+        mean.allocator()->allocate();
+        sd.allocator()->allocate();
+
+        ARM_COMPUTE_EXPECT(!src.info()->is_resizable(), framework::LogLevel::ERRORS);
+        ARM_COMPUTE_EXPECT(!dst.info()->is_resizable(), framework::LogLevel::ERRORS);
+        ARM_COMPUTE_EXPECT(!mean.info()->is_resizable(), framework::LogLevel::ERRORS);
+        ARM_COMPUTE_EXPECT(!sd.info()->is_resizable(), framework::LogLevel::ERRORS);
+
+        // Fill tensors
+        fill(AccessorType(src), AccessorType(mean), AccessorType(sd));
+
+        // Compute function
+        norm.run();
+
+        return dst;
+    }
+
+    SimpleTensor<T> compute_reference(const TensorShape &shape0, const TensorShape &shape1, DataType dt)
+    {
+        // Create reference
+        SimpleTensor<T> ref_src{ shape0, dt, 1 };
+        SimpleTensor<T> ref_mean{ shape1, dt, 1 };
+        SimpleTensor<T> ref_sd{ shape1, dt, 1 };
+
+        // Fill reference
+        fill(ref_src, ref_mean, ref_sd);
+
+        return reference::normalize_planar_yuv_layer(ref_src, ref_mean, ref_sd);
+    }
+
+    TensorType      _target{};
+    SimpleTensor<T> _reference{};
+    DataType        _data_type{};
+};
+
+template <typename TensorType, typename AccessorType, typename FunctionType, typename T>
+class NormalizePlanarYUVLayerValidationFixture : public NormalizePlanarYUVLayerValidationFixedPointFixture<TensorType, AccessorType, FunctionType, T>
+{
+public:
+    template <typename...>
+    void setup(TensorShape shape0, TensorShape shape1, DataType dt)
+    {
+        NormalizePlanarYUVLayerValidationFixedPointFixture<TensorType, AccessorType, FunctionType, T>::setup(shape0, shape1, dt);
+    }
+};
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
+#endif /* ARM_COMPUTE_TEST_NORMALIZE_PLANAR_YUV_LAYER_FIXTURE */
diff --git a/tests/validation/reference/NormalizePlanarYUVLayer.cpp b/tests/validation/reference/NormalizePlanarYUVLayer.cpp
new file mode 100644
index 0000000..2442943
--- /dev/null
+++ b/tests/validation/reference/NormalizePlanarYUVLayer.cpp
@@ -0,0 +1,69 @@
+/*
+ * 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.
+ */
+#include "NormalizePlanarYUVLayer.h"
+
+#include "tests/validation/Helpers.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+namespace reference
+{
+// NormalizePlanarYUV Layer for floating point type
+template <typename T, typename std::enable_if<is_floating_point<T>::value, int>::type *>
+SimpleTensor<T> normalize_planar_yuv_layer(const SimpleTensor<T> &src, const SimpleTensor<T> &mean, const SimpleTensor<T> &sd)
+{
+    SimpleTensor<T> result(src.shape(), src.data_type());
+
+    const auto cols       = static_cast<int>(src.shape()[0]);
+    const auto rows       = static_cast<int>(src.shape()[1]);
+    const auto depth      = static_cast<int>(src.shape()[2]);
+    const int  upper_dims = src.shape().total_size() / (cols * rows * depth);
+
+    for(int r = 0; r < upper_dims; ++r)
+    {
+        for(int i = 0; i < depth; ++i)
+        {
+            for(int k = 0; k < rows; ++k)
+            {
+                for(int l = 0; l < cols; ++l)
+                {
+                    const int pos = l + k * cols + i * rows * cols + r * cols * rows * depth;
+                    result[pos]   = (src[pos] - mean[i]) / sd[i];
+                }
+            }
+        }
+    }
+    return result;
+}
+
+template SimpleTensor<half> normalize_planar_yuv_layer(const SimpleTensor<half> &src, const SimpleTensor<half> &mean, const SimpleTensor<half> &sd);
+
+} // namespace reference
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
diff --git a/tests/validation/reference/NormalizePlanarYUVLayer.h b/tests/validation/reference/NormalizePlanarYUVLayer.h
new file mode 100644
index 0000000..c8740a3
--- /dev/null
+++ b/tests/validation/reference/NormalizePlanarYUVLayer.h
@@ -0,0 +1,44 @@
+/*
+ * 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_NORMALIZE_PLANAR_YUV_LAYER_H__
+#define __ARM_COMPUTE_TEST_NORMALIZE_PLANAR_YUV_LAYER_H__
+
+#include "tests/SimpleTensor.h"
+#include "tests/validation/Helpers.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+namespace reference
+{
+template <typename T, typename std::enable_if<is_floating_point<T>::value, int>::type * = nullptr>
+SimpleTensor<T> normalize_planar_yuv_layer(const SimpleTensor<T> &src, const SimpleTensor<T> &mean, const SimpleTensor<T> &sd);
+} // namespace reference
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
+#endif /* __ARM_COMPUTE_TEST_NORMALIZE_PLANAR_YUV_LAYER_H__ */