How can I use TensorFlow library from matlab environment?

32 Ansichten (letzte 30 Tage)
I want to use or access TensorFlow library from Matlab environment, especially for machine learning. How can I use TensorFlow library from matlab ?
  5 Kommentare
David Willingham
David Willingham am 7 Feb. 2022
Since 2021a, MATLAB supports Framework Interoperability with TensorFlow with a direct importer via importTensorFlowNetwork and importTensorFlowLayers. I'd recommend using these functions over importKerasNetwork or importKerasLayers.
Oluwanisola Ibikunle
Oluwanisola Ibikunle am 25 Jan. 2023
As a follow-up question: If a number of models have been developed in Keras using different versions of Tensorflow (v2.2-2.10) and as a result different versions of Python (3.6,3.9,etc)
How can I build a single MATLAB pipeline that can do a system call from MATLAB such that it abstracts away the different Python versions?
At this point, I'm interested in loading the saved models in .h5 and performing model inference in MATLAB.

Melden Sie sich an, um zu kommentieren.

Akzeptierte Antwort

David Willingham
David Willingham am 27 Aug. 2020
Bearbeitet: David Willingham am 4 Jul. 2021
Hi,
MATLAB supports Framework Interoperability with TensorFlow with a direct importer via importTensorFlowNetwork and importTensorFlowLayers.
Regards,
David Willingham
Deep Learning Product Manager, MathWorks
  1 Kommentar
Jack Ferrari
Jack Ferrari am 31 Mai 2022
On a related note, with R2022a and later, you can run cosimulations and deploy applications using MATLAB or Simulink with TensorFlow Lite (.tflite) models. With TFLite models, an import process is unnecessary. Instead, you can directly load .tflite model files into MATLAB with loadTFLiteModel.
This functionality requires the Deep Learning Toolbox Interface for TensorFlow Lite. More info on requirements can be found here: Prerequisites for Deep Learning with TensorFlow Lite Models.

Melden Sie sich an, um zu kommentieren.

Weitere Antworten (4)

Louis Yu
Louis Yu am 5 Mai 2017
I've went about working on a middle-man solution for new users to Tensorflow that typically utilize Matlab. It's a simple GUI interface that auto-codes the user inputs in the Matlab GUI into a python script that can be run utilizing the Tensorflow Python Interface.
See the file exchange @ Tensorflow for Matlab
For more details on reasons for this implementation, a paper was written over the effort conducted "Benchmarking Open Source Machine Learning Frameworks and Development of Auto-coding Interface for Entry-Level Engineers" .
There is also a github repository I utilized for some test-cases that include Tensorflow snippets: Github Repository for Benchmarking
  2 Kommentare
Kenn
Kenn am 31 Jul. 2017
Bearbeitet: Kenn am 31 Jul. 2017
@Louis Yu, congrats on your effort, Tensorflow for Matlab. It looks like it has high potential but there's something not quite right. I mentioned several bugs I found, which make it look like the code could never have worked. Perhaps an earlier, incomplete version got uploaded? Mind taking a look?
Louis Yu
Louis Yu am 5 Aug. 2017
Updates have been made to the file exchange upload. I've added a few helpful pieces for new users.

Melden Sie sich an, um zu kommentieren.


Tom Anderson
Tom Anderson am 15 Apr. 2017
TensorFlow is a free Python library developed by Google Brain. As of April 2017, it has APIs in other languages (C++, Java and Go), but they are experimental.
MATLAB is a proprietary programming language developed by Mathworks (non-free). It has interfaces to other languages, including Python. Mathworks offers its own Neural Network Toolbox™. There are also plenty of other deep learning MATLAB toolboxes, many listed at http://deeplearning.net/software_links/.
However, running TensorFlow with MATLAB is not supported by TensorFlow or Mathworks. As we can see, the MathWorks File Exchange does not have http://au.mathworks.com/matlabcentral/fileexchange/?utf8=%E2%9C%93&term=tensorflow. Neither does GitHub https://github.com/search?utf8=%E2%9C%93&q=matlab+tensorflow&type=. Until there is a bridge between the two, I would suggest that we are best advised to learn TensorFlow in the supported language (Python). Alternatively, if you already have paid for the MATLAB licenses (Neural Network Toolbox $1,500 Individual) (or belong to an organisation that has), you may consider learning MATLAB and the Neural Network Toolbox. This field is moving quickly, so be prepared to keep paying for the new version of Neural Network Toolbox (updates are not included in the price).
  1 Kommentar
Christopher Stokely
Christopher Stokely am 1 Feb. 2019
Bearbeitet: Christopher Stokely am 3 Feb. 2019
The MATLAB Neural network toolbos for an individual is about $50, not more than $1000. Look at the license options on the MathWorks website for individual license for noncommercial use.
The Rosetta Stone of deep learning is ONNX (Open Neural Network Exchange), which allows model's to be transferred (I think) between environments such as PyTorch, MXNet, Core ML, Caffe2, TensorFlow, Microsoft Cognitive Toolkit, and MATLAB - I think. Check it out for yourself by searching for the command
exportONNXNetwork
References:
1) Open Neural Network Exchange https://github.com/onnx
From Mathworks webpage description, see the attached figure to see a summary of supported deep learning frameworks.

Melden Sie sich an, um zu kommentieren.


Ashiqur Rahman
Ashiqur Rahman am 8 Mär. 2017
Bearbeitet: Ashiqur Rahman am 8 Mär. 2017

Alam Noor
Alam Noor am 12 Feb. 2019

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by