Importazione ed esportazione di Deep Learning

Importare ed esportare reti da/a piattaforme esterne di Deep Learning

Importare reti e grafici di livello da TensorFlow™ 2, TensorFlow-Keras, PyTorch^®, dal formato di modello ONNX™ (Open Neural Network Exchange) e da Caffe. È inoltre possibile esportare reti e grafici di livello Deep Learning Toolbox™ su TensorFlow 2 e nel formato di modello ONNX. Per ulteriori informazioni, vedere Reti neurali deep preaddestrate e Interoperability Between Deep Learning Toolbox, TensorFlow, PyTorch, and ONNX.

Per eseguire le funzioni di importazione ed esportazione in Deep Learning Toolbox è necessario disporre dei pacchetti di supporto. Se il pacchetto di supporto non è installato, ciascuna funzione fornisce un link per il download del pacchetto di supporto corrispondente nell'Add-On Explorer. Si consiglia di scaricare il pacchetto di supporto nella posizione predefinita per la versione MATLAB^® in uso. È inoltre possibile scaricare direttamente i pacchetti di supporto dai seguenti link.

Le funzioni importONNXNetwork, importONNXLayers, importONNXFunction e exportONNXNetwork richiedono il convertitore Deep Learning Toolbox per il formato di modello ONNX. Per scaricare il pacchetto di supporto, andare a https://www.mathworks.com/matlabcentral/fileexchange/67296-deep-learning-toolbox-converter-for-onnx-model-format.
Le funzioni importTensorFlowNetwork, importTensorFlowLayers, exportNetworkToTensorFlow, importKerasNetwork e importKerasLayers richiedono il convertitore Deep Learning Toolbox per i modelli TensorFlow. Per scaricare il pacchetto di supporto, andare a https://www.mathworks.com/matlabcentral/fileexchange/64649-deep-learning-toolbox-converter-for-tensorflow-models.
La funzione importNetworkFromPyTorch richiede il convertitore Deep Learning Toolbox per i modelli PyTorch. Per scaricare il pacchetto di supporto, andare a https://www.mathworks.com/matlabcentral/fileexchange/111925.
Le funzioni importCaffeNetwork e importCaffeLayers richiedono l’importatore Deep Learning Toolbox per i modelli Caffe. Per scaricare il pacchetto di supporto, andare a https://www.mathworks.com/matlabcentral/fileexchange/61735-deep-learning-toolbox-importer-for-caffe-models.

Funzioni

espandi tutto

Importazione della rete

Importazione TensorFlow

`importTensorFlowNetwork`	Import pretrained TensorFlow network
`importTensorFlowLayers`	Import layers from TensorFlow network
`importKerasNetwork`	Import pretrained Keras network and weights
`importKerasLayers`	Import layers from Keras network

Importazione PyTorch

importNetworkFromPyTorch Import PyTorch model as MATLAB network

Importazione Caffe

`importCaffeNetwork`	Import pretrained convolutional neural network models from Caffe
`importCaffeLayers`	Import convolutional neural network layers from Caffe

Importazione ONNX

`importONNXNetwork`	Import pretrained ONNX network
`importONNXLayers`	Import layers from ONNX network
`importONNXFunction`	Import pretrained ONNX network as a function

Parametri importati da `importONNXFunction`

`ONNXParameters`	Parameters of imported ONNX network for deep learning
`freezeParameters`	Convert learnable network parameters in `ONNXParameters` to nonlearnable
`unfreezeParameters`	Convert nonlearnable network parameters in `ONNXParameters` to learnable
`addParameter`	Add parameter to `ONNXParameters` object
`removeParameter`	Remove parameter from `ONNXParameters` object

Assemblaggio della rete

`functionLayer`	Function layer
`findPlaceholderLayers`	Find placeholder layers in network architecture imported from Keras or ONNX
`replaceLayer`	Replace layer in layer graph or network
`assembleNetwork`	Assemble deep learning network from pretrained layers
`PlaceholderLayer`	Layer replacing an unsupported Keras or ONNX layer, or unsupported functionality from `functionToLayerGraph`
`addLayers`	Add layers to layer graph or network
`removeLayers`	Remove layers from layer graph or network

Esportazione della rete

`exportNetworkToTensorFlow`	Export Deep Learning Toolbox network or layer graph to TensorFlow
`exportONNXNetwork`	Export network to ONNX model format

Argomenti

Interoperability Between Deep Learning Toolbox, TensorFlow, PyTorch, and ONNX
Learn how to import networks from TensorFlow, PyTorch, and ONNX and use the imported networks for common Deep Learning Toolbox workflows. Learn how to export networks to TensorFlow and ONNX.
Tips on Importing Models from TensorFlow, PyTorch, and ONNX
Tips on importing Deep Learning Toolbox networks or layer graphs from TensorFlow, PyTorch, and ONNX.
Reti neurali deep preaddestrate
Apprendere come scaricare e utilizzare le reti neurali convoluzionali preaddestrate per la classificazione, il transfer learning e l’estrazione di feature.
Inference Comparison Between TensorFlow and Imported Networks for Image Classification
Perform prediction in TensorFlow with a pretrained network, import the network into MATLAB using importTensorFlowNetwork, and then compare inference results between TensorFlow and MATLAB networks.
Inference Comparison Between ONNX and Imported Networks for Image Classification
Perform prediction in ONNX with a pretrained network, import the network into MATLAB using importONNXNetwork, and then compare inference results between ONNX and MATLAB networks.
Assemble Network from Pretrained Keras Layers
This example shows how to import the layers from a pretrained Keras network, replace the unsupported layers with custom layers, and assemble the layers into a network ready for prediction.
Replace Unsupported Keras Layer with Function Layer
This example shows how to import the layers from a pretrained Keras network, replace the unsupported layers with function layers, and assemble the layers into a network ready for prediction.
Classify Images in Simulink with Imported TensorFlow Network
Import a pretrained TensorFlow network using importTensorFlowNetwork, and then use the Predict block for image classification in Simulink^®.
Deploy Imported TensorFlow Model with MATLAB Compiler
Import third-party pretrained networks and deploy the networks using MATLAB Compiler™.
Select Function to Import ONNX Pretrained Network
Import an ONNX pretrained network using importONNXNetwork, importONNXLayers, or importONNXFunction.