COMPMID-1838: Add 4D softmax support for NEON and achieve parity with CL

Change-Id: I15c4a747cde2536b1caba2baf4ded9ca76e6dae2
Signed-off-by: Manuel Bottini <manuel.bottini@arm.com>
Reviewed-on: https://review.mlplatform.org/487
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: VidhyaSudhan Loganathan <vidhyasudhan.loganathan@arm.com>

1 files changed, 132 insertions, 20 deletions

diff --git a/src/runtime/NEON/functions/NESoftmaxLayer.cpp b/src/runtime/NEON/functions/NESoftmaxLayer.cpp
index 9be9e6817a..36b7d47d28 100644
--- a/src/runtime/NEON/functions/NESoftmaxLayer.cpp
+++ b/src/runtime/NEON/functions/NESoftmaxLayer.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2018 ARM Limited.
+ * Copyright (c) 2017-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -25,54 +25,155 @@
 
 #include "arm_compute/core/Helpers.h"
 #include "arm_compute/core/NEON/kernels/NESoftmaxLayerKernel.h"
+#include "arm_compute/core/utils/misc/ShapeCalculator.h"
 #include "arm_compute/runtime/NEON/NEScheduler.h"
+#include "utils/TypePrinter.h"
 
 #include <cfloat>
 
-using namespace arm_compute;
-
+namespace arm_compute
+{
 NESoftmaxLayer::NESoftmaxLayer(std::shared_ptr<IMemoryManager> memory_manager)
-    : _memory_group(std::move(memory_manager)), _max_kernel(), _softmax_kernel(), _fill_border_kernel(), _max(), _tmp()
+    : _memory_group(std::move(memory_manager)), _max_kernel(), _softmax_kernel(), _flat_or_reshape_kernel_ptr(nullptr), _fill_border_kernel(), _reshape_kernel(), _max(), _tmp(), _input_flattened(),
+      _output_flattened(), _needs_flattening(false)
+{
+}
+
+void NESoftmaxLayer::configure_reshape_input_kernel(const ITensor *input, const ITensor *output, size_t axis)
 {
+    // Flatten the input
+    const TensorShape shape_flatten = misc::shape_calculator::compute_softmax_shape(input->info(), axis);
+
+    // Initialize the flat input
+    _input_flattened.allocator()->init(input->info()->clone()->set_is_resizable(true).reset_padding().set_tensor_shape(shape_flatten));
+
+    // If we need to flatten the input, we can use NEFlattenKernel or NEReshapeKernel
+    // If flattening on the third axes, we use NEFlattenKernel.
+    // In all other cases we have to use NEReshapeKernel
+    if(axis != 3)
+    {
+        auto reshape_kernel_ptr = support::cpp14::make_unique<NEReshapeLayerKernel>();
+        reshape_kernel_ptr->configure(input, &_input_flattened);
+        _flat_or_reshape_kernel_ptr = std::move(reshape_kernel_ptr);
+    }
+    else
+    {
+        auto flatten_kernel_ptr = support::cpp14::make_unique<NEFlattenLayerKernel>();
+        flatten_kernel_ptr->configure(input, &_input_flattened);
+        _flat_or_reshape_kernel_ptr = std::move(flatten_kernel_ptr);
+    }
+
+    // We need to init the output tensor here. Indeed, the reshape kernel expects
+    // both tensors to be already initialized
+    auto_init_if_empty(*output->info(), *input->info()->clone());
 }
 
 void NESoftmaxLayer::configure(ITensor *input, ITensor *output, float beta, size_t axis)
 {
+    // Perform validation step
     ARM_COMPUTE_ERROR_ON_NULLPTR(input, output);
-    ARM_COMPUTE_UNUSED(axis);
+    ARM_COMPUTE_ERROR_THROW_ON(NESoftmaxLayer::validate(input->info(), output->info(), beta, axis));
 
-    // Configure Kernels
-    _max_kernel.configure(input, &_max);
-    _fill_border_kernel.configure(input, _max_kernel.border_size(), BorderMode::REPLICATE);
-    _softmax_kernel.configure(input, &_max, output, beta, &_tmp);
+    // We don't need flattening only in the case the input is 2D and axis is 1
+    _needs_flattening = axis != 1;
+
+    // If we are dealing with a 4D tensor, we will:
+    // - Flatten the input, so that we end up with a [width*height*depth] * batches 2D tensor
+    // - Execute all the pipeline (reduction + normalization) on the flattened tensor
+    // - Reshape the flattened output into the real output
+    if(_needs_flattening)
+    {
+        // Add to the memory manager _input_flattened
+        _memory_group.manage(&_input_flattened);
+
+        // Configure  _flatten_kernel and _input_flattened
+        configure_reshape_input_kernel(input, output, axis);
+    }
+
+    // We want to deal with a 2D input. Either it is the flattened version of the original input (4D case)
+    // or it is the original input case (2D case)
+    ITensor *input_2D = (_needs_flattening ? &_input_flattened : input);
+
+    // Create intermediate tensors shapes
+    const TensorInfo input_info    = input_2D->info()->clone()->reset_padding().set_is_resizable(true);
+    DataType         tmp_data_type = is_data_type_quantized_asymmetric(input_2D->info()->data_type()) ? DataType::F32 : input_2D->info()->data_type();
+    TensorInfo       tensor_info_tmp(input_info.clone()->set_data_type(tmp_data_type));
 
     // Init intermediate tensors
-    _max.allocator()->init(*_max.info());
-    _tmp.allocator()->init(*_tmp.info());
+    TensorShape max_sum_shape = input_2D->info()->tensor_shape();
+    max_sum_shape.set(0, 1);
+    _max.allocator()->init(input_info.clone()->set_tensor_shape(max_sum_shape));
+    _tmp.allocator()->init(tensor_info_tmp);
 
     // Manage intermediate buffers
     _memory_group.manage(&_max);
     _memory_group.manage(&_tmp);
 
-    // Allocate intermediate tensors
+    // Configure Kernels
+    _max_kernel.configure(input_2D, &_max);
+    if(_needs_flattening)
+    {
+        // Add to the memory manager _output_flattened
+        _memory_group.manage(&_output_flattened);
+
+        // The normalization kernel stores the result in a flat output tensor
+        _softmax_kernel.configure(input_2D, &_max, &_output_flattened, beta, &_tmp);
+        _input_flattened.allocator()->allocate();
+
+        // Reshape the flat output into the requested (4D) output
+        _reshape_kernel.configure(&_output_flattened, output);
+
+        // Allocate the intermediate flat tensors
+        _output_flattened.allocator()->allocate();
+    }
+    else
+    {
+        // Softmax 2D case
+        _fill_border_kernel.configure(input_2D, _max_kernel.border_size(), BorderMode::REPLICATE);
+        _softmax_kernel.configure(input_2D, &_max, output, beta, &_tmp);
+    }
+
+    // Allocate intermediate buffers
     _max.allocator()->allocate();
     _tmp.allocator()->allocate();
 }
 
 Status NESoftmaxLayer::validate(const ITensorInfo *input, const ITensorInfo *output, float beta, size_t axis)
 {
-    ARM_COMPUTE_RETURN_ERROR_ON_MSG(axis != 1, "Axis must be 1 for NEON");
-
     // Perform validation step
     ARM_COMPUTE_RETURN_ERROR_ON_NULLPTR(input, output);
-    ARM_COMPUTE_RETURN_ERROR_ON_MSG(input->num_dimensions() > 2, "Only 2D inputs are supported");
-
-    const TensorShape max_shape           = TensorShape(input->tensor_shape()).set(0, 1);
-    const TensorInfo  tensor_info_max_sum = TensorInfo(*input).set_tensor_shape(max_shape).reset_padding();
-    const TensorInfo  dont_care;
+    ARM_COMPUTE_RETURN_ERROR_ON_MSG(input->num_dimensions() > 4, "Only up to 4 dimensions are supported");
+    ARM_COMPUTE_UNUSED(beta);
+    ARM_COMPUTE_RETURN_ERROR_ON(axis < 1 || input->num_dimensions() < axis);
+
+    // Create intermediate tensor info
+    DataType         tmp_data_type = input->data_type();
+    const TensorInfo tensor_info_tmp(input->clone()->set_data_type(tmp_data_type).set_is_resizable(true));
+
+    TensorShape max_sum_shape = input->tensor_shape();
+    max_sum_shape.set(0, 1);
+    const TensorInfo tensor_info_max_sum(input->clone()->set_tensor_shape(max_sum_shape).set_data_type(tmp_data_type).set_quantization_info(input->quantization_info()).set_is_resizable(true));
+    const TensorInfo dont_care;
+
+    const bool needs_flattening = (axis != 1);
+
+    if(needs_flattening)
+    {
+        const TensorShape shape_flatten = misc::shape_calculator::compute_softmax_shape(input, axis);
+        TensorInfo        tensor_info_flat(input->clone()->set_tensor_shape(shape_flatten).set_is_resizable(true));
+
+        if(axis != 3)
+        {
+            ARM_COMPUTE_RETURN_ON_ERROR(NEReshapeLayerKernel::validate(input, &tensor_info_flat));
+        }
+        else
+        {
+            ARM_COMPUTE_RETURN_ON_ERROR(NEFlattenLayerKernel::validate(input, &tensor_info_flat));
+        }
+    }
 
     ARM_COMPUTE_RETURN_ON_ERROR(NELogits1DMaxKernel::validate(input, &tensor_info_max_sum));
-    ARM_COMPUTE_RETURN_ON_ERROR(NELogits1DSoftmaxKernel::validate(input, &tensor_info_max_sum, output, beta, &dont_care));
+    ARM_COMPUTE_RETURN_ON_ERROR(NELogits1DSoftmaxKernel::validate(&tensor_info_tmp, &tensor_info_max_sum, output, beta, &dont_care));
 
     return Status{};
 }
@@ -81,9 +182,20 @@ void NESoftmaxLayer::run()
 {
     _memory_group.acquire();
 
+    if(_needs_flattening)
+    {
+        NEScheduler::get().schedule(_flat_or_reshape_kernel_ptr.get(), Window::DimY);
+    }
+
     NEScheduler::get().schedule(&_fill_border_kernel, Window::DimY);
     NEScheduler::get().schedule(&_max_kernel, Window::DimY);
     NEScheduler::get().schedule(&_softmax_kernel, Window::DimY);
 
+    if(_needs_flattening)
+    {
+        NEScheduler::get().schedule(&_reshape_kernel, Window::DimY);
+    }
+
     _memory_group.release();
 }
+} // namespace arm_compute
\ No newline at end of file