diff --git a/Android.bp b/Android.bp
index 957c8e2..950192c 100644
--- a/Android.bp
+++ b/Android.bp
@@ -468,6 +468,9 @@
         "src/cpu/kernels/instancenorm/generic/neon/impl.cpp",
         "src/cpu/kernels/internal/CpuDepthwiseConv2dAssemblyWrapperKernel.cpp",
         "src/cpu/kernels/internal/CpuPool2dAssemblyWrapperKernel.cpp",
+        "src/cpu/kernels/l2normlayer/generic/neon/fp16.cpp",
+        "src/cpu/kernels/l2normlayer/generic/neon/fp32.cpp",
+        "src/cpu/kernels/l2normlayer/generic/neon/impl.cpp",
         "src/cpu/kernels/maxunpool/generic/neon/fp16.cpp",
         "src/cpu/kernels/maxunpool/generic/neon/fp32.cpp",
         "src/cpu/kernels/maxunpool/generic/neon/impl.cpp",
diff --git a/filelist.json b/filelist.json
index 88d98ae..6e28635 100644
--- a/filelist.json
+++ b/filelist.json
@@ -1578,7 +1578,12 @@
           "common": [
             "src/core/NEON/kernels/NEL2NormalizeLayerKernel.cpp",
             "src/runtime/NEON/functions/NEL2NormalizeLayer.cpp"
-          ]
+          ],
+          "neon":{
+            "common":["src/cpu/kernels/l2normlayer/generic/neon/impl.cpp"],
+            "fp32":["src/cpu/kernels/l2normlayer/generic/neon/fp32.cpp"],
+            "fp16":["src/cpu/kernels/l2normlayer/generic/neon/fp16.cpp"]
+          }
         }
       },
       "Logical": {
diff --git a/src/core/NEON/kernels/NEL2NormalizeLayerKernel.cpp b/src/core/NEON/kernels/NEL2NormalizeLayerKernel.cpp
index 9bda82d..8ab0288 100644
--- a/src/core/NEON/kernels/NEL2NormalizeLayerKernel.cpp
+++ b/src/core/NEON/kernels/NEL2NormalizeLayerKernel.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2021 Arm Limited.
+ * Copyright (c) 2017-2022 Arm Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -30,11 +30,13 @@
 #include "arm_compute/core/Utils.h"
 #include "arm_compute/core/Validate.h"
 #include "arm_compute/core/Window.h"
+#include "src/common/cpuinfo/CpuIsaInfo.h"
 #include "src/core/NEON/NEMath.h"
+#include "src/core/common/Registrars.h"
 #include "src/core/helpers/AutoConfiguration.h"
 #include "src/core/helpers/WindowHelpers.h"
+#include "src/cpu/kernels/l2normlayer/list.h"
 
-#include "src/core/NEON/wrapper/wrapper.h"
 #include <arm_neon.h>
 #include <cmath>
 
@@ -44,90 +46,64 @@
 {
 constexpr int max_input_tensor_dim = 3;
 
-template <typename T, int S>
-void l2_normalize_X(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window)
+struct L2NormalizeLayerSelectorData
 {
-    using ExactTagType = typename wrapper::traits::neon_vector<T, S>::tag_type;
+    DataType            dt;
+    unsigned int        actual_axis;
+    cpuinfo::CpuIsaInfo isa;
+};
 
-    const int  window_step_x  = 16 / data_size_from_type(in->info()->data_type());
-    const auto window_start_x = static_cast<int>(window.x().start());
-    const auto window_end_x   = static_cast<int>(window.x().end());
+using L2NormalizeLayerKernelSelctorPtr = std::add_pointer<bool(const L2NormalizeLayerSelectorData &data)>::type;
 
-    Window win_collapsed = window.collapse_if_possible(window, Window::DimZ);
-    win_collapsed.set(Window::DimX, Window::Dimension(0, 1, 1));
+using L2NormalizeLayerPtr = std::add_pointer<void(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window, size_t axis)>::type;
 
-    Iterator input_it(in, win_collapsed);
-    Iterator sum_it(sum, win_collapsed);
-    Iterator output_it(out, win_collapsed);
-
-    execute_window_loop(win_collapsed, [&](const Coordinates &)
-    {
-        const auto in_ptr  = reinterpret_cast<const T *>(input_it.ptr());
-        const auto out_ptr = reinterpret_cast<T *>(output_it.ptr());
-
-        const T    sum_value      = *reinterpret_cast<const T *>(sum_it.ptr());
-        const T    norm_value     = static_cast<T>(1.f) / std::sqrt(std::max(sum_value, static_cast<T>(epsilon)));
-        const auto vec_norm_value = wrapper::vdup_n(norm_value, ExactTagType{});
-
-        // Compute elements over vector steps
-        int x = window_start_x;
-        for(; x <= (window_end_x - window_step_x); x += window_step_x)
-        {
-            wrapper::vstore(out_ptr + x, wrapper::vmul(wrapper::vloadq(in_ptr + x), vec_norm_value));
-        }
-
-        // Compute left-over elements
-        for(; x < window_end_x; ++x)
-        {
-            out_ptr[x] = in_ptr[x] * norm_value;
-        }
-    },
-    input_it, sum_it, output_it);
-}
-
-template <typename T, int S>
-void l2_normalize_YZ(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window, size_t axis)
+struct L2NormalizeLayerKernel
 {
-    using ExactTagType = typename wrapper::traits::neon_vector<T, S>::tag_type;
+    const char                            *name;
+    const L2NormalizeLayerKernelSelctorPtr is_selected;
+    L2NormalizeLayerPtr                    ukernel;
+};
 
-    const int  window_step_x  = 16 / data_size_from_type(in->info()->data_type());
-    const auto window_start_x = static_cast<int>(window.x().start());
-    const auto window_end_x   = static_cast<int>(window.x().end());
-
-    Window win = window;
-    win.set(Window::DimX, Window::Dimension(0, 1, 1));
-
-    Window window_sum(win);
-    window_sum.set(axis, Window::Dimension(0, 0, 0));
-
-    Iterator input_it(in, win);
-    Iterator sum_it(sum, window_sum);
-    Iterator output_it(out, win);
-
-    const auto vec_eps = wrapper::vdup_n(static_cast<T>(epsilon), ExactTagType{});
-
-    execute_window_loop(win, [&](const Coordinates &)
+static const L2NormalizeLayerKernel available_kernels[] =
+{
     {
-        const auto in_ptr  = reinterpret_cast<const T *>(input_it.ptr());
-        const auto sum_ptr = reinterpret_cast<const T *>(sum_it.ptr());
-        const auto out_ptr = reinterpret_cast<T *>(output_it.ptr());
-
-        // Compute elements over vector steps
-        int x = window_start_x;
-        for(; x <= (window_end_x - window_step_x); x += window_step_x)
-        {
-            const auto vec_norm_value = wrapper::vinvsqrt(wrapper::vmax(wrapper::vloadq(sum_ptr + x), vec_eps));
-            wrapper::vstore(out_ptr + x, wrapper::vmul(wrapper::vloadq(in_ptr + x), vec_norm_value));
-        }
-
-        // Compute left-over elements
-        for(; x < window_end_x; ++x)
-        {
-            const T norm_value = static_cast<T>(1.f) / std::sqrt(std::max(sum_ptr[x], static_cast<T>(epsilon)));
-            out_ptr[x]         = in_ptr[x] * norm_value;
-        }
+        "fp32_neon_l2normalize_x",
+        [](const L2NormalizeLayerSelectorData & data) { return data.dt == DataType::F32 && data.actual_axis == Window::DimX; },
+        REGISTER_FP32_NEON(arm_compute::cpu::neon_fp32_l2_normalize_x)
     },
-    input_it, sum_it, output_it);
+    {
+        "fp32_neon_l2normalize_yz",
+        [](const L2NormalizeLayerSelectorData & data) { return data.dt == DataType::F32 && data.actual_axis != Window::DimX; },
+        REGISTER_FP32_NEON(arm_compute::cpu::neon_fp32_l2_normalize_yz)
+    },
+    {
+        "fp16_neon_l2normalize_x",
+        [](const L2NormalizeLayerSelectorData & data) { return data.dt == DataType::F16 && data.isa.fp16 && data.actual_axis == Window::DimX; },
+        REGISTER_FP16_NEON(arm_compute::cpu::neon_fp16_l2_normalize_x),
+    },
+    {
+        "fp16_neon_l2normalize_yz",
+        [](const L2NormalizeLayerSelectorData & data) { return data.dt == DataType::F16 && data.isa.fp16 && data.actual_axis != Window::DimX; },
+        REGISTER_FP16_NEON(arm_compute::cpu::neon_fp16_l2_normalize_yz),
+    },
+};
+
+/** Micro-kernel selector
+ *
+ * @param[in] data Selection data passed to help pick the appropriate micro-kernel
+ *
+ * @return A matching micro-kernel else nullptr
+ */
+const L2NormalizeLayerKernel *get_implementation(const L2NormalizeLayerSelectorData &data)
+{
+    for(const auto &uk : available_kernels)
+    {
+        if(uk.is_selected(data))
+        {
+            return &uk;
+        }
+    }
+    return nullptr;
 }
 
 Status validate_arguments(const ITensorInfo *input, const ITensorInfo *sum, const ITensorInfo *output, int axis, float epsilon)
@@ -212,18 +188,10 @@
         ARM_COMPUTE_ERROR("Unsupported normalization axis");
     }
 
-    switch(_input->info()->data_type())
-    {
-        case DataType::F32:
-            (_actual_axis == Window::DimX) ? l2_normalize_X<float, 4>(_input, _sum, _output, _epsilon, window) : l2_normalize_YZ<float, 4>(_input, _sum, _output, _epsilon, window, _actual_axis);
-            break;
-#ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC
-        case DataType::F16:
-            (_actual_axis == Window::DimX) ? l2_normalize_X<float16_t, 8>(_input, _sum, _output, _epsilon, window) : l2_normalize_YZ<float16_t, 8>(_input, _sum, _output, _epsilon, window, _actual_axis);
-            break;
-#endif // __ARM_FEATURE_FP16_VECTOR_ARITHMETIC
-        default:
-            ARM_COMPUTE_ERROR("Not implemented");
-    }
+    const auto *uk = get_implementation(L2NormalizeLayerSelectorData{ _output->info()->data_type(), _actual_axis, CPUInfo::get().get_isa() });
+    ARM_COMPUTE_ERROR_ON(uk == nullptr);
+    ARM_COMPUTE_ERROR_ON(uk->ukernel == nullptr);
+
+    uk->ukernel(_input, _sum, _output, _epsilon, window, _actual_axis);
 }
 } // namespace arm_compute
diff --git a/src/cpu/kernels/l2normlayer/generic/neon/fp16.cpp b/src/cpu/kernels/l2normlayer/generic/neon/fp16.cpp
new file mode 100644
index 0000000..ed84c10
--- /dev/null
+++ b/src/cpu/kernels/l2normlayer/generic/neon/fp16.cpp
@@ -0,0 +1,45 @@
+/*
+ * Copyright (c) 2022 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.
+ */
+#if defined(__ARM_FEATURE_FP16_VECTOR_ARITHMETIC) && defined(ENABLE_FP16_KERNELS)
+
+#include "src/cpu/kernels/l2normlayer/generic/neon/impl.h"
+
+#include "arm_compute/core/Helpers.h"
+namespace arm_compute
+{
+namespace cpu
+{
+void neon_fp16_l2_normalize_x(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window, size_t unused_axis)
+{
+    ARM_COMPUTE_UNUSED(unused_axis);
+    return l2_normalize_x<float16_t, 8>(in, sum, out, epsilon, window);
+}
+
+void neon_fp16_l2_normalize_yz(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window, size_t axis)
+{
+    return l2_normalize_yz<float16_t, 8>(in, sum, out, epsilon, window, axis);
+}
+} // namespace cpu
+} // namespace arm_compute
+#endif /* defined(__ARM_FEATURE_FP16_VECTOR_ARITHMETIC) && defined(ENABLE_FP16_KERNELS) */
diff --git a/src/cpu/kernels/l2normlayer/generic/neon/fp32.cpp b/src/cpu/kernels/l2normlayer/generic/neon/fp32.cpp
new file mode 100644
index 0000000..be32bdc
--- /dev/null
+++ b/src/cpu/kernels/l2normlayer/generic/neon/fp32.cpp
@@ -0,0 +1,45 @@
+/*
+ * Copyright (c) 2022 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 "src/cpu/kernels/l2normlayer/generic/neon/impl.h"
+
+#include "arm_compute/core/Helpers.h"
+
+namespace arm_compute
+{
+namespace cpu
+{
+void neon_fp32_l2_normalize_x(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window, size_t unused_axis)
+{
+    ARM_COMPUTE_UNUSED(unused_axis);
+    return l2_normalize_x<float, 4>(in, sum, out, epsilon, window);
+}
+
+void neon_fp32_l2_normalize_yz(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window, size_t axis)
+{
+    return l2_normalize_yz<float, 4>(in, sum, out, epsilon, window, axis);
+}
+
+} // namespace cpu
+} // namespace arm_compute
diff --git a/src/cpu/kernels/l2normlayer/generic/neon/impl.cpp b/src/cpu/kernels/l2normlayer/generic/neon/impl.cpp
new file mode 100644
index 0000000..2886537
--- /dev/null
+++ b/src/cpu/kernels/l2normlayer/generic/neon/impl.cpp
@@ -0,0 +1,131 @@
+/*
+ * Copyright (c) 2017-2022 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 "src/cpu/kernels/l2normlayer/generic/neon/impl.h"
+
+#include "arm_compute/core/Helpers.h"
+#include "arm_compute/core/TensorInfo.h"
+#include "src/core/NEON/wrapper/wrapper.h"
+#include "src/core/common/Registrars.h"
+
+#include <cstddef>
+
+namespace arm_compute
+{
+namespace cpu
+{
+template <typename T, int S>
+void l2_normalize_x(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window)
+{
+    using ExactTagType = typename wrapper::traits::neon_vector<T, S>::tag_type;
+
+    const int  window_step_x  = 16 / data_size_from_type(in->info()->data_type());
+    const auto window_start_x = static_cast<int>(window.x().start());
+    const auto window_end_x   = static_cast<int>(window.x().end());
+
+    Window win_collapsed = window.collapse_if_possible(window, Window::DimZ);
+    win_collapsed.set(Window::DimX, Window::Dimension(0, 1, 1));
+
+    Iterator input_it(in, win_collapsed);
+    Iterator sum_it(sum, win_collapsed);
+    Iterator output_it(out, win_collapsed);
+
+    execute_window_loop(win_collapsed, [&](const Coordinates &)
+    {
+        const auto in_ptr  = reinterpret_cast<const T *>(input_it.ptr());
+        const auto out_ptr = reinterpret_cast<T *>(output_it.ptr());
+
+        const T    sum_value      = *reinterpret_cast<const T *>(sum_it.ptr());
+        const T    norm_value     = static_cast<T>(1.f) / std::sqrt(std::max(sum_value, static_cast<T>(epsilon)));
+        const auto vec_norm_value = wrapper::vdup_n(norm_value, ExactTagType{});
+
+        // Compute elements over vector steps
+        int x = window_start_x;
+        for(; x <= (window_end_x - window_step_x); x += window_step_x)
+        {
+            wrapper::vstore(out_ptr + x, wrapper::vmul(wrapper::vloadq(in_ptr + x), vec_norm_value));
+        }
+
+        // Compute left-over elements
+        for(; x < window_end_x; ++x)
+        {
+            out_ptr[x] = in_ptr[x] * norm_value;
+        }
+    },
+    input_it, sum_it, output_it);
+}
+
+template <typename T, int S>
+void l2_normalize_yz(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window, size_t axis)
+{
+    using ExactTagType = typename wrapper::traits::neon_vector<T, S>::tag_type;
+
+    const int  window_step_x  = 16 / data_size_from_type(in->info()->data_type());
+    const auto window_start_x = static_cast<int>(window.x().start());
+    const auto window_end_x   = static_cast<int>(window.x().end());
+
+    Window win = window;
+    win.set(Window::DimX, Window::Dimension(0, 1, 1));
+
+    Window window_sum(win);
+    window_sum.set(axis, Window::Dimension(0, 0, 0));
+
+    Iterator input_it(in, win);
+    Iterator sum_it(sum, window_sum);
+    Iterator output_it(out, win);
+
+    const auto vec_eps = wrapper::vdup_n(static_cast<T>(epsilon), ExactTagType{});
+
+    execute_window_loop(win, [&](const Coordinates &)
+    {
+        const auto in_ptr  = reinterpret_cast<const T *>(input_it.ptr());
+        const auto sum_ptr = reinterpret_cast<const T *>(sum_it.ptr());
+        const auto out_ptr = reinterpret_cast<T *>(output_it.ptr());
+
+        // Compute elements over vector steps
+        int x = window_start_x;
+        for(; x <= (window_end_x - window_step_x); x += window_step_x)
+        {
+            const auto vec_norm_value = wrapper::vinvsqrt(wrapper::vmax(wrapper::vloadq(sum_ptr + x), vec_eps));
+            wrapper::vstore(out_ptr + x, wrapper::vmul(wrapper::vloadq(in_ptr + x), vec_norm_value));
+        }
+
+        // Compute left-over elements
+        for(; x < window_end_x; ++x)
+        {
+            const T norm_value = static_cast<T>(1.f) / std::sqrt(std::max(sum_ptr[x], static_cast<T>(epsilon)));
+            out_ptr[x]         = in_ptr[x] * norm_value;
+        }
+    },
+    input_it, sum_it, output_it);
+}
+
+template void l2_normalize_yz<float, 4>(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window, size_t axis);
+template void l2_normalize_x<float, 4>(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window);
+
+#if defined(__ARM_FEATURE_FP16_VECTOR_ARITHMETIC) && defined(ENABLE_FP16_KERNELS)
+template void l2_normalize_yz<float16_t, 8>(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window, size_t axis);
+template void l2_normalize_x<float16_t, 8>(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window);
+#endif //defined(__ARM_FEATURE_FP16_VECTOR_ARITHMETIC) && defined(ENABLE_FP16_KERNELS)
+} // namespace cpu
+} // namespace arm_compute
diff --git a/src/cpu/kernels/l2normlayer/generic/neon/impl.h b/src/cpu/kernels/l2normlayer/generic/neon/impl.h
new file mode 100644
index 0000000..98391fb
--- /dev/null
+++ b/src/cpu/kernels/l2normlayer/generic/neon/impl.h
@@ -0,0 +1,44 @@
+/*
+ * Copyright (c) 2022 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 SRC_CORE_NEON_KERNELS_L2NORMLAYER_LIST_H
+#define SRC_CORE_NEON_KERNELS_L2NORMLAYER_LIST_H
+
+#include <cstddef>
+
+namespace arm_compute
+{
+class ITensor;
+class Window;
+
+namespace cpu
+{
+template <typename T, int S>
+void l2_normalize_x(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window);
+
+template <typename T, int S>
+void l2_normalize_yz(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window, size_t axis);
+
+} // namespace cpu
+} // namespace arm_compute
+#endif //SRC_CORE_NEON_KERNELS_L2NORMLAYER_LIST_H
diff --git a/src/cpu/kernels/l2normlayer/list.h b/src/cpu/kernels/l2normlayer/list.h
new file mode 100644
index 0000000..2bad7f5
--- /dev/null
+++ b/src/cpu/kernels/l2normlayer/list.h
@@ -0,0 +1,41 @@
+/*
+ * Copyright (c) 2022 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 SRC_CORE_NEON_KERNELS_L2NORMLAYER_LIST_H
+#define SRC_CORE_NEON_KERNELS_L2NORMLAYER_LIST_H
+namespace arm_compute
+{
+namespace cpu
+{
+#define DECLARE_L2NORMLAYER_KERNEL(func_name) \
+    void func_name(const ITensor *in, const ITensor *sum, ITensor *out, float epsilon, const Window &window, size_t axis)
+
+DECLARE_L2NORMLAYER_KERNEL(neon_fp16_l2_normalize_x);
+DECLARE_L2NORMLAYER_KERNEL(neon_fp16_l2_normalize_yz);
+DECLARE_L2NORMLAYER_KERNEL(neon_fp32_l2_normalize_x);
+DECLARE_L2NORMLAYER_KERNEL(neon_fp32_l2_normalize_yz);
+
+#undef DECLARE_L2NORMLAYER_KERNEL
+} // namespace cpu
+} // namespace arm_compute
+#endif //SRC_CORE_NEON_KERNELS_L2NORMLAYER_LIST_H