openvino/docs/OV_Runtime_UG/ShapeInference.md

Changing Input Shapes

@sphinxdirective

.. toctree:: :maxdepth: 1 :hidden:

troubleshooting_reshape_errors

OpenVINO™ enables you to change model input shape during the application runtime. It may be useful when you want to feed the model an input that has different size than the model input shape. The following instructions are for cases where you need to change the model input shape repeatedly.

.. note::

If you need to do this only once, prepare a model with updated shapes via :doc:model conversion API <openvino_docs_MO_DG_Deep_Learning_Model_Optimizer_DevGuide>. For more information, refer to the :ref:Specifying input_shape Parameter <when_to_specify_input_shapes> article.

The reshape method ++++++++++++++++++++

The reshape method is used as ov::Model::reshape in C++ and Model.reshape <api/ie_python_api/_autosummary/openvino.runtime.Model.html#openvino.runtime.Model.reshape>__ in Python. The method updates input shapes and propagates them down to the outputs of the model through all intermediate layers. The code below is an example of how to set a new batch size with the reshape method:

.. tab-set::

.. tab-item:: C++ :sync: cpp

  .. doxygensnippet:: docs/snippets/ShapeInference.cpp
     :language: cpp
     :fragment: picture_snippet

.. tab-item:: Python :sync: py

  .. doxygensnippet:: docs/snippets/ShapeInference.py
     :language: Python
     :fragment: picture_snippet

The diagram below presents the results of using the method, where the size of model input is changed with an image input:

.. image:: _static/images/original_vs_reshaped_model.svg

When using the reshape method, you may take one of the approaches:

.. _usage_of_reshape_method:

1. You can pass a new shape to the method in order to change the input shape of the model with a single input. See the example of adjusting spatial dimensions to the input image:

   .. tab-set::

   .. tab-item:: C++ :sync: cpp

     .. doxygensnippet:: docs/snippets/ShapeInference.cpp
        :language: cpp
        :fragment: spatial_reshape
   

   .. tab-item:: Python :sync: py

     .. doxygensnippet:: docs/snippets/ShapeInference.py
        :language: python
        :fragment: simple_spatials_change
   

   To do the opposite - to resize input image to match the input shapes of the model, use the :doc:pre-processing API <openvino_docs_OV_UG_Preprocessing_Overview>.

2. You can express a reshape plan, specifying the input by the port, the index, and the tensor name:

   .. tab-set::

   .. tab-item:: Port

     .. tab-set::
   
        .. tab-item:: C++
           :sync: cpp
   
           ``map<ov::Output<ov::Node>, ov::PartialShape`` specifies input by passing actual input port:
   
           .. doxygensnippet:: docs/snippets/ShapeInference.cpp
              :language: cpp
              :fragment: [obj_to_shape]
   
        .. tab-item:: Python
           :sync: py
   
           ``openvino.runtime.Output`` dictionary key specifies input by passing actual input object.
           Dictionary values representing new shapes could be ``PartialShape``:
   
           .. doxygensnippet:: docs/snippets/ShapeInference.py
              :language: python
              :fragment: [obj_to_shape]
   

   .. tab-item:: Index

     .. tab-set::
   
        .. tab-item:: C++
           :sync: cpp
   
           ``map<size_t, ov::PartialShape>`` specifies input by its index:
   
           .. doxygensnippet:: docs/snippets/ShapeInference.cpp
              :language: cpp
              :fragment: [idx_to_shape]
   
        .. tab-item:: Python
           :sync: py
   
           ``int`` dictionary key specifies input by its index.
           Dictionary values representing new shapes could be ``tuple``:
   
           .. doxygensnippet:: docs/snippets/ShapeInference.py
              :language: python
              :fragment: [idx_to_shape]
   

   .. tab-item:: Tensor Name

     .. tab-set::
   
        .. tab-item:: C++
           :sync: cpp
   
           ``map<string, ov::PartialShape>`` specifies input by its name:
   
           .. doxygensnippet:: docs/snippets/ShapeInference.cpp
              :language: cpp
              :fragment: [name_to_shape]
   
        .. tab-item:: Python
           :sync: py
   
           ``str`` dictionary key specifies input by its name.
           Dictionary values representing new shapes could be ``str``:
   
           .. doxygensnippet:: docs/snippets/ShapeInference.py
              :language: python
              :fragment: [name_to_shape]
   

You can find the usage scenarios of the reshape method in :doc:Hello Reshape SSD Samples <openvino_inference_engine_samples_hello_reshape_ssd_README>.

.. note::

In some cases, models may not be ready to be reshaped. Therefore, a new input shape cannot be set neither with :doc:Model Optimizer <openvino_docs_MO_DG_Deep_Learning_Model_Optimizer_DevGuide> nor the reshape method.

The set_batch method ++++++++++++++++++++

The meaning of the model batch may vary depending on the model design. To change the batch dimension of the model, :ref:set the layout <declare_model_s_layout> and call the set_batch method.

.. tab-set::

.. tab-item:: C++ :sync: cpp

  .. doxygensnippet:: docs/snippets/ShapeInference.cpp
     :language: cpp
     :fragment: set_batch

.. tab-item:: Python :sync: py

  .. doxygensnippet:: docs/snippets/ShapeInference.py
     :language: Python
     :fragment: set_batch

The set_batch method is a high-level API of the reshape functionality, so all information about the reshape method implications are applicable for set_batch too, including the troubleshooting section.

Once you set the input shape of the model, call the compile_model method to get a CompiledModel object for inference with updated shapes.

There are other approaches to change model input shapes during the stage of :ref:IR generation <when_to_specify_input_shapes> or :ref:model representation <openvino_docs_OV_UG_Model_Representation> in OpenVINO Runtime.

.. important::

Shape-changing functionality could be used to turn dynamic model input into a static one and vice versa. Always set static shapes when the shape of data is NOT going to change from one inference to another. Setting static shapes can avoid memory and runtime overheads for dynamic shapes which may vary depending on hardware plugin and model used. For more information, refer to the :doc:Dynamic Shapes <openvino_docs_OV_UG_DynamicShapes>.

Additional Resources ####################

• :doc:Extensibility documentation <openvino_docs_Extensibility_UG_Intro> - describes a special mechanism in OpenVINO that allows adding support of shape inference for custom operations.
• ov::Model::reshape <classov_1_1Model.html#doxid-classov-1-1-model-1aa21aff80598d5089d591888a4c7f33ae>__ - in OpenVINO Runtime C++ API
• Model.reshape <api/ie_python_api/_autosummary/openvino.runtime.Model.html#openvino.runtime.Model.reshape>__ - in OpenVINO Runtime Python API.
• :doc:Dynamic Shapes <openvino_docs_OV_UG_DynamicShapes>
• :doc:OpenVINO samples <openvino_docs_OV_UG_Samples_Overview>
• :doc:Preprocessing API <openvino_docs_OV_UG_Preprocessing_Overview>

@endsphinxdirective