- Global InferenceEngine::Blob::buffer () noexcept=0
- Cast to MemoryBlob and use new wlock/rwlock API instead. Blob class can represent compound blob, which do not refer to the only solid memory.
- Global InferenceEngine::Blob::cbuffer () const noexcept=0
- Cast to MemoryBlob and use new MemoryBlob::rmap() function instead. Blob class can represent compound blob, which do not refer to the only solid memory.
- Global InferenceEngine::Blob::element_size () const noexcept=0
- Cast to MemoryBlob and use its API instead. Blob class can represent compound blob, which do not refer to the only solid memory.
- Global InferenceEngine::Blob::product (const SizeVector &dims) noexcept
- Cast to MemoryBlob and use its API instead.
- Class InferenceEngine::CNNNetReader
- Use InferenceEngine::Core::ReadNetwork methods. This API will be removed in 2021.1
- Global InferenceEngine::CNNNetwork::AddExtension (InferenceEngine::IShapeInferExtensionPtr extension)
- Use Core::AddExtension to add an extension to the library
- Global InferenceEngine::CNNNetwork::begin () const
- Use CNNNetwork::getFunction() and work with ngraph::Function directly. The method will be removed in 2021.1
- Global InferenceEngine::CNNNetwork::end () const
- Use CNNNetwork::getFunction() and work with ngraph::Function directly. The method will be removed in 2021.1
- Global InferenceEngine::CNNNetwork::getLayerByName (const char *layerName) const
- Migrate to IR v10 and work with ngraph::Function directly. The method will be removed in 2021.1
- Global InferenceEngine::CNNNetwork::getPrecision () const
- Network precision does not make sence, use precision on egdes. The method will be removed in 2021.1
- Global InferenceEngine::CNNNetwork::size () const
- Use CNNNetwork::layerCount() instead. The method will be removed in 2021.1
- Global InferenceEngine::Core::SetLogCallback (IErrorListener &listener) const
- IErrorListener is not used anymore. An exception is thrown in case of any unexpected situations. The function will be removed in 2021.1 release.
- Global InferenceEngine::CreateShapeInferExtension (IShapeInferExtension *&ext, ResponseDesc *resp) noexcept
- Migrate to IR v10 and implement shape inference in the ngraph::op::Op::validate_and_infer_types method This API will be removed in 2021.1 release.
- Global InferenceEngine::Data::Data (const std::string &name, const SizeVector &a_dims, Precision _precision, Layout layout=NCHW)
- Use Data(const std::string &name, const TensorDesc& desc)
- Global InferenceEngine::Data::getCreatorLayer ()
- Migrate to IR v10 and work with ngraph::Function directly. The method will be removed in 2021.1
- Global InferenceEngine::Data::getInputTo ()
- Migrate to IR v10 and work with ngraph::Function directly. The method will be removed in 2021.1
- Global InferenceEngine::ExecutableNetwork::GetMappedTopology (std::map< std::string, std::vector< PrimitiveInfo::Ptr >> &deployedTopology)
- Use ExecutableNetwork::GetExecGraphInfo to get information about an internal graph. This method will be removed in 2021.1 release.
- Global InferenceEngine::Extension::getFactoryFor (ILayerImplFactory *&factory, const CNNLayer *cnnLayer, ResponseDesc *resp) noexcept override
- Use IExtension::getImplementation to get a concrete implementation. The method will be removed in 2021.1 release.
- Global InferenceEngine::Extension::getPrimitiveTypes (char **&types, unsigned int &size, ResponseDesc *resp) noexcept override
- Use IExtension::getImplTypes to get implementation types for a particular node. The method will removed in 2021.1 release.
- Global InferenceEngine::Extension::getShapeInferImpl (IShapeInferImpl::Ptr &impl, const char *type, ResponseDesc *resp) noexcept override
- Implement ngraph::op::Op::validate_and_infer_types method in a custom ngraph operation The method will be removed in 2021.1 release.
- Global InferenceEngine::Extension::getShapeInferTypes (char **&types, unsigned int &size, ResponseDesc *resp) noexcept override
- Implement ngraph::op::Op::validate_and_infer_types method in a custom ngraph operation The method will be removed in 2021.1 release.
- Global InferenceEngine::Extension::SetLogCallback (InferenceEngine::IErrorListener &listener) noexcept override
- IErrorListener is not used anymore. StatusCode is provided in case of unexpected situations The method will be removed in 2021.1 release.
- Class InferenceEngine::ICNNNetReader
- Use InferenceEngine::Core::ReadNetwork methods. This API will be removed in 2021.1
- Global InferenceEngine::ICNNNetwork::AddExtension (const IShapeInferExtensionPtr &extension, ResponseDesc *resp) noexcept
- Use Core::AddExtension to add an extension to the library
- Global InferenceEngine::ICNNNetwork::addLayer (const CNNLayerPtr &layer) noexcept=0
- Migrate to IR v10 and work with ngraph::Function directly. The method will be removed in 2021.1
- Global InferenceEngine::ICNNNetwork::getData (const char *dname) noexcept=0
- Migrate to IR v10 and work with ngraph::Function directly. The method will be removed in 2021.1
- Global InferenceEngine::ICNNNetwork::getLayerByName (const char *layerName, CNNLayerPtr &out, ResponseDesc *resp) const noexcept=0
- Migrate to IR v10 and work with ngraph::Function directly. The method will be removed in 2021.1
- Global InferenceEngine::ICNNNetwork::getName (char *pName, size_t len) const noexcept=0
- Use ICNNNetwork::getName() instead. The method will be removed in 2021.1
- Global InferenceEngine::ICNNNetwork::getPrecision () const noexcept=0
- Network precision does not make sence, use precision on egdes. The method will be removed in 2021.1
- Global InferenceEngine::ICNNNetwork::getStats (ICNNNetworkStats **stats, ResponseDesc *resp) const noexcept
- Migrate to IR v10 and use quantization approach with FakeQuantize
- Class InferenceEngine::ICNNNetworkStats
Migrate to IR v10 and use quantization approach with FakeQuantize
- Class InferenceEngine::IErrorListener
- IErrorListener is not used anymore. An exception is thrown / StatusCode set in case of any unexpected situations The class will be removed in 2021.1 release.
- Global InferenceEngine::IExecutableNetwork::GetMappedTopology (std::map< std::string, std::vector< PrimitiveInfo::Ptr >> &deployedTopology, ResponseDesc *resp) noexcept=0
- Use ExecutableNetwork::GetExecGraphInfo to get information about an internal graph. The method will be removed in 2021.1 release.
- Global InferenceEngine::IExtension::getFactoryFor (ILayerImplFactory *&factory, const CNNLayer *cnnLayer, ResponseDesc *resp) noexcept
- Use IExtension::getImplementation to get a concrete implementation The method will be removed in 2021.1 release.
- Global InferenceEngine::IExtension::getPrimitiveTypes (char **&types, unsigned int &size, ResponseDesc *resp) noexcept
- Use IExtension::getImplTypes to get implementation types for a particular node The method will be removed in 2021.1 release.
- Class InferenceEngine::IInferencePlugin
- Use InferenceEngine::Core instead. Will be removed in 2021.1
- Global InferenceEngine::IInferencePlugin::SetLogCallback (IErrorListener &listener) noexcept=0
- IErrorListener is not used anymore. StatusCode is provided in case of unexpected situations This API will be removed in 2021.1 release.
- Class InferenceEngine::ILayerImplFactory
Implement IExtension::getImplTypes and IExtension::getImplementation The interface will be removed in 2021.1 release.
- Class InferenceEngine::InferencePlugin
- Use InferenceEngine::Core instead. Will be removed in 2021.1
- Class InferenceEngine::IShapeInferExtension
Implement a custom ngraph operation derived from ngraph::op::Op in IExtension implementation
- Global InferenceEngine::IShapeInferExtension::getShapeInferImpl (IShapeInferImpl::Ptr &impl, const char *type, ResponseDesc *resp) noexcept=0
- Implement ngraph::op::Op::validate_and_infer_types method in a custom ngraph operation. The method will be removed in 2021.1 release.
- Global InferenceEngine::IShapeInferExtension::getShapeInferTypes (char **&types, unsigned int &size, ResponseDesc *resp) noexcept=0
- Implement ngraph::op::Op::validate_and_infer_types method in a custom ngraph operation. The method will be removed in 2021.1 release.
- Global InferenceEngine::IShapeInferExtension::SetLogCallback (InferenceEngine::IErrorListener &listener) noexcept
- IErrorListener is not used anymore. StatusCode is provided in case of unexpected situations The method will be removed in 2021.1 release.
- Global InferenceEngine::IShapeInferExtensionPtr
- Migrate to IR v10 and implement shape inference in the ngraph::op::Op::validate_and_infer_types method This API will be removed in 2021.1 release.
- Class InferenceEngine::IShapeInferImpl
Implement ngraph::op::Op::validate_and_infer_types method in a custom ngraph operation. The interface will be removed in 2021.1 release.
- Global InferenceEngine::make_so_pointer (const file_name_t &name)
- Use make_so_pointer with IExtension as template argument type.
- Global InferenceEngine::make_so_pointer (const file_name_t &name)
- Use make_so_pointer with IExtension as template argument type.
- Global InferenceEngine::MemoryBlob::buffer () noexcept override=0
- Use wmap() or rwmap() API instead.
- Global InferenceEngine::MemoryBlob::cbuffer () const noexcept override=0
- Use rmap() function instead.
- Class InferenceEngine::NetworkNodeStats
Migrate to IR v10 and use quantization approach with FakeQuantize
- Class InferenceEngine::PluginDispatcher
- Use InferenceEngine::Core instead. Will be removed in 2021.1
- Global InferenceEngine::PluginDispatcher::getPluginByDevice (const std::string &deviceName) const
- Use InferenceEngine::Core to work with devices by name
- Class InferenceEngine::PrimitiveInfo
- Use ExecutableNetwork::GetExecGraphInfo to get information about an internal graph. This structure will be removed in 2021.1 release.
- Class InferenceEngine::ShapeInferExtension
- Use a common Extension class. The interface will be removed in 2021.1 release.
- Class InferenceEngine::TensorInfo
Use ExecutableNetwork::GetExecGraphInfo to get information about an internal graph. This API will be removed in 2021.1 release.