diff options
| author | Pablo Tello <pablo.tello@arm.com> | 2019-03-04 14:14:02 +0000 |
|---|---|---|
| committer | Gian Marco Iodice <gianmarco.iodice@arm.com> | 2019-03-19 10:18:10 +0000 |
| commit | 3dd5b6884a65c06bcb9d15589ee2dc2978e3b336 (patch) | |
| tree | e45ccae66b69c8db853ac883080c1c6358a57aec /src/runtime/NEON/functions/NEWidthConcatenateLayer.cpp | |
| parent | 2f7c149f36fa3e6296aba6de666962947f032558 (diff) | |
| download | ComputeLibrary-3dd5b6884a65c06bcb9d15589ee2dc2978e3b336.tar.gz | |
COMPMID-1933: Implement NEHeightConcatenateLayer.
Added support to concactenate tensors along the Y axis in NEConcatenateLayer.
Change-Id: Ib714bfcf9954cc35918efa7d52fc9164bb08bdf6
Signed-off-by: Pablo Tello <pablo.tello@arm.com>
Reviewed-on: https://review.mlplatform.org/c/841
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/runtime/NEON/functions/NEWidthConcatenateLayer.cpp')
| -rw-r--r-- | src/runtime/NEON/functions/NEWidthConcatenateLayer.cpp | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/src/runtime/NEON/functions/NEWidthConcatenateLayer.cpp b/src/runtime/NEON/functions/NEWidthConcatenateLayer.cpp index 17c352b8f3..9fce13cbd7 100644 --- a/src/runtime/NEON/functions/NEWidthConcatenateLayer.cpp +++ b/src/runtime/NEON/functions/NEWidthConcatenateLayer.cpp @@ -48,7 +48,7 @@ inline Status NEWidthConcatenateLayer::validate_internal(const std::vector<Tenso // Output auto inizialitation if not yet initialized TensorInfo tmp_output_info = *output->clone(); - TensorShape output_shape = arm_compute::misc::shape_calculator::calculate_width_concatenate_shape(inputs_vector); + TensorShape output_shape = arm_compute::misc::shape_calculator::calculate_concatenate_shape(inputs_vector, Window::DimX); auto_init_if_empty(tmp_output_info, output_shape, 1, inputs_vector[0]->data_type()); unsigned int width_offset = 0; @@ -71,7 +71,7 @@ inline void NEWidthConcatenateLayer::configure_internal(std::vector<TensorType * { inputs_vector_info.emplace_back(inputs_vector.at(i)->info()); } - TensorShape output_shape = arm_compute::misc::shape_calculator::calculate_width_concatenate_shape(inputs_vector); + TensorShape output_shape = arm_compute::misc::shape_calculator::calculate_concatenate_shape(inputs_vector, Window::DimX); // Output auto inizialitation if not yet initialized auto_init_if_empty(*output->info(), output_shape, 1, inputs_vector[0]->info()->data_type()); |