How to work with Tensorflow on Android platform?

Question

Google has made the TENSORFLOW open source for developers..

Is there any way out for using this on android?

The link is here TensorFlow.

I would love to have some directions to work on this API.

Political

Answer 1

The TensorFlow source repository includes an Android example application, with some documentation.

The Android example includes a pre-trained model for image classification, and uses this to classify images captured by the camera. Typically you would build and train a model using the Python API; generate a serialised version of the model as a GraphDef protocol buffer (and possibly a checkpoint of the model parameters); and then load that and run inference steps using the C++ API.

Scrutiny

thank you @mrry.. Is this API like TESSERACT, in terms of building, compiling and running it using C++ (NDK)? – Political

Is is necessary to work with Python API, to train my android model of tensorflow? Because I don't know coding in python platform. – Political

I'm not familiar with Tesseract, so can't speak to any comparisons. Also it isn't strictly necessary to train your model using the Python API (as the C++ API is also available) but the majority of libraries for TensorFlow (gradients, RNNs, etc.) are currently written in Python, so that API is often easier to use. – Scrutiny

Thank you for better deep involvement, but being new to android, how can I train python with android? How can I make setup of it? Same as NDK or any other way is there to train it!? – Political

Answer 2

The TensorFlow source repository includes an Android example application, Repo

The Android example includes a pre-trained model for image classification, and uses this to classify images captured by the camera. Typically you would build and train a model using the Python API; generate a serialised version of the model as a MobileNetV2

Titfer