aboutsummaryrefslogtreecommitdiff
path: root/delegate/src/Convolution.hpp
diff options
context:
space:
mode:
Diffstat (limited to 'delegate/src/Convolution.hpp')
-rw-r--r--delegate/src/Convolution.hpp92
1 files changed, 68 insertions, 24 deletions
diff --git a/delegate/src/Convolution.hpp b/delegate/src/Convolution.hpp
index 17189027dc..a8559e2548 100644
--- a/delegate/src/Convolution.hpp
+++ b/delegate/src/Convolution.hpp
@@ -115,7 +115,7 @@ TfLiteStatus VisitConv2dOperator(DelegateData& delegateData,
}
- armnn::TensorInfo filterTensorInfo = GetTensorInfoForTfLiteTensor(tfLiteFilterTensor);
+ const armnn::TensorInfo& filterTensorInfo = GetTensorInfoForTfLiteTensor(tfLiteFilterTensor);
armnn::TensorInfo biasTensorInfo;
if(biasEnabled)
@@ -181,12 +181,11 @@ TfLiteStatus VisitConv2dOperator(DelegateData& delegateData,
armnn::IConnectableLayer* layer = delegateData.m_Network->AddConvolution2dLayer(descriptor);
layer->SetBackendId(setBackend);
- if(tflite::IsConstantTensor(&tfLiteContext->tensors[tfLiteNode->inputs->data[1]]))
+ if(filterTensorInfo.IsConstant())
{
auto filter =
CreateConstTensor(&tfLiteContext->tensors[tfLiteNode->inputs->data[1]],
- filterTensorInfo,
- armnn::Optional<armnn::PermutationVector &>());
+ filterTensorInfo);
armnn::IConnectableLayer *weightsLayer = delegateData.m_Network->AddConstantLayer(filter);
weightsLayer->GetOutputSlot(0).Connect(layer->GetInputSlot(1u));
@@ -196,7 +195,7 @@ TfLiteStatus VisitConv2dOperator(DelegateData& delegateData,
if (biasEnabled)
{
const TfLiteTensor& tfLiteBiasTensor = tfLiteTensors[tfLiteNode->inputs->data[2]];
- if(tflite::IsConstantTensor(&tfLiteBiasTensor))
+ if(biasTensorInfo.IsConstant())
{
auto biasTensor = CreateConstTensor(&tfLiteBiasTensor, biasTensorInfo);
armnn::IConnectableLayer* biasLayer = delegateData.m_Network->AddConstantLayer(biasTensor);
@@ -206,6 +205,18 @@ TfLiteStatus VisitConv2dOperator(DelegateData& delegateData,
}
}
+ // The data input can also be constant, so we must check that this is also allocated to an input slot
+ if(inputTensorInfo.IsConstant())
+ {
+ auto input =
+ CreateConstTensor(&tfLiteContext->tensors[tfLiteNode->inputs->data[0]],
+ inputTensorInfo);
+
+ armnn::IConnectableLayer *inputLayer = delegateData.m_Network->AddConstantLayer(input);
+ inputLayer->GetOutputSlot(0).Connect(layer->GetInputSlot(0u));
+ inputLayer->GetOutputSlot(0).SetTensorInfo(inputTensorInfo);
+ }
+
ARMNN_ASSERT(layer != nullptr);
armnn::IOutputSlot& outputSlot = layer->GetOutputSlot(0);
@@ -290,7 +301,7 @@ TfLiteStatus VisitConv3dOperator(DelegateData& delegateData,
}
- armnn::TensorInfo filterTensorInfo = GetTensorInfoForTfLiteTensor(tfLiteFilterTensor);
+ const armnn::TensorInfo& filterTensorInfo = GetTensorInfoForTfLiteTensor(tfLiteFilterTensor);
armnn::TensorInfo biasTensorInfo;
if(biasEnabled)
@@ -354,11 +365,10 @@ TfLiteStatus VisitConv3dOperator(DelegateData& delegateData,
// Add a constant layer for weights and biases if inputs are constant,
// which are connected to the Convolution3d layer as inputs.
- if (tflite::IsConstantTensor(&tfLiteFilterTensor))
+ if (filterTensorInfo.IsConstant())
{
auto filter = CreateConstTensor(&tfLiteFilterTensor,
- filterTensorInfo,
- armnn::Optional<armnn::PermutationVector&>());
+ filterTensorInfo);
armnn::IConnectableLayer* weightsLayer = delegateData.m_Network->AddConstantLayer(filter);
ARMNN_ASSERT(weightsLayer != nullptr);
@@ -370,11 +380,10 @@ TfLiteStatus VisitConv3dOperator(DelegateData& delegateData,
if(biasEnabled)
{
const TfLiteTensor& tfLiteBiasTensor = tfLiteTensors[tfLiteNode->inputs->data[2]];
- if(tflite::IsConstantTensor(&tfLiteBiasTensor))
+ if(biasTensorInfo.IsConstant())
{
auto biases = CreateConstTensor(&tfLiteBiasTensor,
- biasTensorInfo,
- armnn::Optional<armnn::PermutationVector&>());
+ biasTensorInfo);
armnn::IConnectableLayer* biasLayer = delegateData.m_Network->AddConstantLayer(biases);
ARMNN_ASSERT(biasLayer != nullptr);
@@ -384,6 +393,18 @@ TfLiteStatus VisitConv3dOperator(DelegateData& delegateData,
}
}
+ // The data input can also be constant, so we must check that this is also allocated to an input slot
+ if(inputTensorInfo.IsConstant())
+ {
+ auto input =
+ CreateConstTensor(&tfLiteContext->tensors[tfLiteNode->inputs->data[0]],
+ inputTensorInfo);
+
+ armnn::IConnectableLayer *inputLayer = delegateData.m_Network->AddConstantLayer(input);
+ inputLayer->GetOutputSlot(0).Connect(layer->GetInputSlot(0u));
+ inputLayer->GetOutputSlot(0).SetTensorInfo(inputTensorInfo);
+ }
+
armnn::IOutputSlot& outputSlot = layer->GetOutputSlot(0);
outputSlot.SetTensorInfo(outputTensorInfo);
@@ -499,7 +520,7 @@ TfLiteStatus VisitDepthwiseConv2dOperator(DelegateData& delegateData,
}
- armnn::TensorInfo filterTensorInfo = GetTensorInfoForTfLiteTensor(tfLiteFilterTensor);
+ const armnn::TensorInfo& filterTensorInfo = GetTensorInfoForTfLiteTensor(tfLiteFilterTensor);
// Assuming input is NHWC
unsigned int inputHeight = inputTensorInfo.GetShape()[1];
@@ -563,7 +584,7 @@ TfLiteStatus VisitDepthwiseConv2dOperator(DelegateData& delegateData,
armnn::IConnectableLayer* layer = delegateData.m_Network->AddDepthwiseConvolution2dLayer(descriptor);
layer->SetBackendId(setBackend);
- if(tflite::IsConstantTensor(&tfLiteFilterTensor))
+ if(filterTensorInfo.IsConstant())
{
// For depthwise the weights layout is the same as for tflite [1, H, W, I*M]. No permutation required.
auto filter = CreateConstTensor(&tfLiteFilterTensor, filterTensorInfo);
@@ -576,7 +597,7 @@ TfLiteStatus VisitDepthwiseConv2dOperator(DelegateData& delegateData,
if (biasEnabled)
{
const TfLiteTensor& tfLiteBiasTensor = tfLiteTensors[tfLiteNode->inputs->data[2]];
- if(tflite::IsConstantTensor(&tfLiteBiasTensor))
+ if(biasTensorInfo.IsConstant())
{
auto biasTensor = CreateConstTensor(&tfLiteBiasTensor, biasTensorInfo);
armnn::IConnectableLayer* biasLayer = delegateData.m_Network->AddConstantLayer(biasTensor);
@@ -586,6 +607,18 @@ TfLiteStatus VisitDepthwiseConv2dOperator(DelegateData& delegateData,
}
}
+ // The data input can also be constant, so we must check that this is also allocated to an input slot
+ if(inputTensorInfo.IsConstant())
+ {
+ auto input =
+ CreateConstTensor(&tfLiteContext->tensors[tfLiteNode->inputs->data[0]],
+ inputTensorInfo);
+
+ armnn::IConnectableLayer *inputLayer = delegateData.m_Network->AddConstantLayer(input);
+ inputLayer->GetOutputSlot(0).Connect(layer->GetInputSlot(0u));
+ inputLayer->GetOutputSlot(0).SetTensorInfo(inputTensorInfo);
+ }
+
ARMNN_ASSERT(layer != nullptr);
armnn::IOutputSlot& outputSlot = layer->GetOutputSlot(0);
@@ -636,19 +669,19 @@ TfLiteStatus VisitTransposeConv2dOperator(DelegateData& delegateData,
return kTfLiteError;
}
- armnn::TensorInfo tensorInfo = GetTensorInfoForTfLiteTensor(tfLiteOutputShapeTensor);
- std::vector<int32_t> outputShape(tensorInfo.GetNumElements());
- if (tensorInfo.GetDataType() == armnn::DataType::Signed32)
+ const armnn::TensorInfo outputShapeTensorInfo = GetTensorInfoForTfLiteTensor(tfLiteOutputShapeTensor);
+ std::vector<int32_t> outputShape(outputShapeTensorInfo.GetNumElements());
+ if (outputShapeTensorInfo.GetDataType() == armnn::DataType::Signed32)
{
- for(unsigned int i=0; i < tensorInfo.GetNumElements(); i++)
+ for(unsigned int i=0; i < outputShapeTensorInfo.GetNumElements(); i++)
{
outputShape[i] = ::tflite::GetTensorData<int32_t>(&tfLiteOutputShapeTensor)[i];
}
}
- if (tensorInfo.GetDataType() == armnn::DataType::QAsymmU8)
+ if (outputShapeTensorInfo.GetDataType() == armnn::DataType::QAsymmU8)
{
- for(unsigned int i=0; i < tensorInfo.GetNumElements(); i++)
+ for(unsigned int i=0; i < outputShapeTensorInfo.GetNumElements(); i++)
{
outputShape[i] = ::tflite::GetTensorData<uint8_t>(&tfLiteOutputShapeTensor)[i];
}
@@ -716,7 +749,7 @@ TfLiteStatus VisitTransposeConv2dOperator(DelegateData& delegateData,
const armnn::TensorInfo& inputTensorInfo = GetTensorInfoForTfLiteTensor(tfLiteInputTensor);
const armnn::TensorInfo& outputTensorInfo = GetTensorInfoForTfLiteTensor(tfLiteOutputTensor, true);
- armnn::TensorInfo filterTensorInfo = GetTensorInfoForTfLiteTensor(tfLiteFilterTensor);
+ const armnn::TensorInfo& filterTensorInfo = GetTensorInfoForTfLiteTensor(tfLiteFilterTensor);
// TfLite uses NHWC tensors
const unsigned int inputHeight = inputTensorInfo.GetShape()[1];
@@ -743,8 +776,7 @@ TfLiteStatus VisitTransposeConv2dOperator(DelegateData& delegateData,
// Set up filter
auto filterTensor = CreateConstTensor(&tfLiteFilterTensor,
- filterTensorInfo,
- armnn::Optional<armnn::PermutationVector&>());
+ filterTensorInfo);
armnn::BackendId setBackend;
if (!delegateData.m_Network)
{
@@ -769,6 +801,18 @@ TfLiteStatus VisitTransposeConv2dOperator(DelegateData& delegateData,
layer->SetBackendId(setBackend);
ARMNN_ASSERT(layer != nullptr);
+ // The data input can be constant, so we must check that this is allocated to an input slot
+ if(inputTensorInfo.IsConstant())
+ {
+ auto input =
+ CreateConstTensor(&tfLiteContext->tensors[tfLiteNode->inputs->data[2]],
+ inputTensorInfo);
+
+ armnn::IConnectableLayer *inputLayer = delegateData.m_Network->AddConstantLayer(input);
+ inputLayer->GetOutputSlot(0).Connect(layer->GetInputSlot(0u));
+ inputLayer->GetOutputSlot(0).SetTensorInfo(inputTensorInfo);
+ }
+
armnn::IOutputSlot& outputSlot = layer->GetOutputSlot(0);
outputSlot.SetTensorInfo(outputTensorInfo);
generated by cgit v1.2.1 (git 2.23.0) at 2024-06-05 20:34:01 +0000