Installing Keras with TensorFlow backend

A few months ago I demonstrated how to install the Keras deep learning library with a Theano backend.

In today’s blog post I provide detailed, step-by-step instructions to install Keras using a TensorFlow backend, originally developed by the researchers and engineers on the Google Brain Team.

I’ll also (optionally) demonstrate how you can integrate OpenCV into this setup for a full-fledged computer vision + deep learning development environment.

To learn more, just keep reading.

Installing Keras with TensorFlow backend

The first part of this blog post provides a short discussion of Keras backends and why we should (or should not) care which one we are using.

From there I provide detailed instructions that you can use to install Keras with a TensorFlow backend for machine learning on your own system.

TensorFlow? Theano? Who cares?

It’s important to start this discussion by saying that Keras is simply a wrapper around more complex numerical computation engines such as TensorFlow and Theano.

Keras abstracts away much of the complexity of building a deep neural network, leaving us with a very simple, nice, and easy to use interface to rapidly build, test, and deploy deep learning architectures.

When it comes to Keras you have two choices for a backend engine — either TensorFlow or Theano. Theano is older than TensorFlow and was originally the only choice when selecting a backend for Keras.

So why might you want to use TensorFlow over Theano?

The short version is that TensorFlow is extremely flexible, allowing you to deploy network computation to multiple CPUs, GPUs, servers, or even mobile systems without having to change a single line of code.

This makes TensorFlow an excellent choice for training distributed deep learning networks in an architecture agnostic way, something that Theano does not (currently) provide.

To be totally honest with you, I started using Keras well before TensorFlow was released (or even rumored to exist) — this was back when Theano was the only possible choice of backend.

I haven’t given much thought to whether Theano or TensorFlow should be my “go to” backend. Theano was working well for what I needed it for, so why bother switching?

My eyes started to open when I ran this recent poll on Twitter asking my followers which backend they preferred when using Keras:

Figure 1: I polled my Twitter followers (@pyimagesearch) to determine whether they preferred using Theano or TensorFlow as their Keras backend.

67% of respondents said they were using TensorFlow as their backend. I was honestly quite surprised. How, as a long-time Keras user, could I possibly be in the minority?

This 67% of respondents might be swayed since TensorFlow is now the default backend when installing Keras…or it could be because many of my followers are finding TensorFlow a better, more efficient backend (and using more TensorFlow specific features).

Regardless of the exact reasoning, there is one thing you cannot dispute: TensorFlow is here to stay.

If you need further proof all you need to do is take a look at this deep learning GitHub activity analysis from François Chollet (creator and maintainer of Keras):

Figure 2: TensorFlow tops the charts as the deep learning library with most GitHub activity.  Keras follows at #2 with Theano all the way at #9.

As we can see, TensorFlow is topping the charts by a mile (#1) with Theano at #9.

While Keras makes it simple for us to switch backends (all we need to do is install our respective backends and edit a simple JSON configuration file), we still need to be mindful of what the trends are telling us: that TensorFlow will continue to be the preferred Keras backend in the (near) future.

Step #1: Setup Python virtual environment (optional)

If you’ve ever read any of my previous tutorials (whether for OpenCV, CUDA, Keras, etc.) you’ve likely picked up on the fact that I’m a huge fan of using Python virtual environments.

I especially recommend Python virtual environments when working in the deep learning ecosystem. Many Python-based deep learning libraries require different versions of various dependencies.

For example, if you wanted to use Keras + Theano together you would need the latest version of Theano (i.e., their latest GitHub commit, which isn’t always the version published on PyPI).

However, if you wanted to try a library such as scikit-theano you would need a previous version of Theano that is not compatible with Keras.

The dependency version issue only compounds as you start to add in other Python libraries, especially deep learning ones (such as TensorFlow), which are volatile in their very nature (since deep learning is a new, fast moving field with updates and new features being pushed online every day).

The solution?

Use Python virtual environments.

I won’t go into a huge rant on the benefits of Python virtual environments (as I’ve already done that in the first half of this blog post), but the gist is that by using Python virtual environments you can create a separate, sequestered Python environment for each of your projects, ensuring that each Python environment is independent of each other. Doing this allows you to totally and completely avoid the version dependency issue.

I’m going to assume that you have both virtualenv and virtualenvwrapper installed on your system (if not, both are pip-installable and require only a small update to your shell configuration; just follow the links above for more information).

Once you have virtualenv and virtualenvwrapper installed, let’s create a Python virtual environment exclusively for our Keras + TensorFlow-based projects:

$ mkvirtualenv keras_tf

I’ll name this virtual environment

keras_tf

  for Keras + TensorFlow (I also have a virtual environment named

keras_th

  for Keras + Theano).

Anytime you need to access a given Python virtual environment just use the

workon

  command:

$ workon <virtual_env_name>

In this particular case we can access the

keras_tf

  virtual environment using:

$ workon keras_tf

Again, while this is an optional step, I really encourage you to take the time to properly setup your development environment if you are even remotely serious about doing deep learning research, experiments, or development using the Python programming language — while it’s more work upfront, you’ll thank me in the long run.

Step #2: Install TensorFlow

Installing TensorFlow is trivially easy as

pip

  will do all the heavy lifting for us.

In fact, the only “real work” we need to do is head over to the TensorFlow “Download and Setup” page and determine which version of TensorFlow we need.

Specifically, you’ll want to scroll down to the “Pip installation” section (skip the actual steps involving installing

pip

  itself) and look for the correct binary to install.

You’ll see a list similar to the screenshot below:

Figure 3: A list of the possible TensorFlow + GPU/CPU + Operating System combinations.

Look through this list and find the TensorFlow binary that matches your particular development environment. Take special note regarding the GPU enabled TensorFlow binaries as many of them require additional dependencies such as the CUDA Toolkit and cuDNN.

Note: I am not covering installing GPU drivers (such as CUDA, cudNN, etc.) in this particular blog post. I’m making the assumption that you already have CUDA, cuDNN, or lack-thereof configured and installed. If this is your first time working with deep learning simply stick to the CPU-only versions of TensorFlow and switch to the GPU later when you are more comfortable with the setup process.

For example, since I am using my Mac OSX machine running Python 2.7 which does not have GPU support enabled (i.e., I need the CPU-only version), I would expert the

TF_BINARY_URL

  environment variable to be:

$ export TF_BINARY_URL=http://ift.tt/2fyKxnK

From there,

pip

  can take take of installing TensorFlow (and any of its dependencies):

$ pip install --upgrade $TF_BINARY_URL

Below you can see a screenshot of TensorFlow being downloaded and installed:

Figure 4: Installing TensorFlow for deep learning via pip into my Python virtual environment.

Assuming your TensorFlow install exited without error you can now test the installation by opening a Python shell and trying to import the

tensorflow

  package:

$ python
>>> import tensorflow
>>>

Step #3: Install Keras

Installing Keras is even easier than installing TensorFlow.

First, let’s install a few Python dependencies:

$ pip install numpy scipy
$ pip install scikit-learn
$ pip install pillow
$ pip install h5py

Followed by installing

keras

  itself:

$ pip install keras

That’s it! Keras is now installed on your system!

Step #4: Verify that your keras.json file is configured correctly

Before we get too far we should check the contents of our

keras.json

  configuration file. You can find this file in

~/.keras/keras.json

 .

Open it using your favorite text editor and take a peak at the contents. The default values should look something like this:

{
    "image_dim_ordering": "tf", 
    "epsilon": 1e-07, 
    "floatx": "float32", 
    "backend": "tensorflow"
}

Specifically, you’ll want to ensure that

image_dim_ordering

  is set to

tf

  (indicating that the TensorFlow image dimension ordering is used rather than

th

  for Theano).

You’ll also want to ensure that the

backend

  is properly set to

tensorflow

  (rather than

theano

 ). Again, both of these requirements should be satisfied by the default Keras configuration but it doesn’t hurt to double check.

Make any required updates (if any) to your configuration file and then exit your editor.

A quick note on image_dim_ordering

You might be wondering what exactly

image_dim_ordering

  controls.

Using TensorFlow, images are represented as NumPy arrays with the shape (height, width, depth), where the depth is the number of channels in the image.

However, if you are using Theano, images are instead assumed to be represented as (depth, height, width).

This little nuance is the source of a lot of headaches when using Keras (and a lot of

if

  statments looking for these particular configurations).

If you are getting strange results when using Keras (or an error message related to the shape of a given tensor) you should:

1. Check your backend.
2. Ensure your image dimesnion ordering matches your backend.

Step #5: Sym-link in OpenCV (optional)

This step is entirely optional, but if you have OpenCV installed on your system and would like to access your OpenCV bindings from a Python virtual environment, you first need to sym-link in the

cv2.so

  file to the

site-packages

  directory of your environment.

To do this, first find where your

cv2.so

  bindings are located on your system:

$ cd /
$ sudo find . -name '*cv2.so*'
./Users/adrianrosebrock/.virtualenvs/cv/lib/python2.7/site-packages/cv2.so
./Users/adrianrosebrock/.virtualenvs/gurus/lib/python2.7/site-packages/cv2.so
./Users/adrianrosebrock/.virtualenvs/keras_th/lib/python2.7/site-packages/cv2.so
./usr/local/lib/python2.7/site-packages/cv2.so

You’ll want to look for the global install of OpenCV which is normally in the

/usr/local/lib

  directory if you built OpenCV from source (unless you specified a custom OpenCV install directory).

Note: The other

cv2.so

  files returned by my

find

  command are simply sym-links back to the original

cv2.so

  file in

/usr/local/lib

 .

From there, change directory into the

site-packages

  directory of your Python virtual environment (in this case the

keras_tf

  environment) and create the sym-link:

$ cd ~/.virtualenvs/keras_tf/lib/python2.7/site-packages/
$ ln -s /usr/local/lib/python2.7/site-packages/cv2.so cv2.so
$ cd ~

Again, this step is totally optional and only needs to be done if you want to have access to OpenCV from inside the

keras_tf

  virtual environment.

Step #6: Test out the Keras + TensorFlow installation

To verify that Keras + TensorFlow have been installed, simply access the

keras_tf

  environment using the

workon

  command, open up a Python shell, and import

keras

 :

Figure 5: Installing and testing our Keras + TensorFlow backend for deep learning.

Specifically, you can see the text

Using TensorFlow backend

  display when importing Keras — this successfully demonstrates that Keras has been installed with the TensorFlow backend.

Provided you performed the optional Step #5 and want to to test out your OpenCV sym-link, try importing your OpenCV bindings as well:

Figure 6: Importing OpenCV and Keras together for deep learning with Python.

If you get an error message related to OpenCV not being found then you’ll want to double check your sym-link and ensure it is pointing to a valid file.

What’s next?

Figure 7: Learn how to utilize Deep Learning and Convolutional Neural Networks to classify the contents of images inside the PyImageSearch Gurus course.

Now that you have Keras + TensorFlow for deep learning installed on your system the obvious question is:

“What’s next?”

I’m glad you asked.

I would suggest:

• Starting with this tutorial on creating a simple neural network with Python and Keras.
• From there, try training your first Convolutional Neural Network (LeNet) for handwritten digit recognition.
• You might also be interested in trying this example on using pre-trained CNN architectures on the ImageNet dataset to recognize 1,000 different common object categories.

Finally, I would also recommend taking a look at the PyImageSearch Gurus course.

Inside the course I have over 168 lessons covering 2,161+ pages of content on Deep Learning, Convolutional Neural Networks, Automatic License Plate Recognition (ANPR), face recognition, and much more!

To learn more about the PyImageSearch Gurus course (and grab 10 FREE sample lessons), just click the button below:

Click here to learn more about PyImageSearch Gurus!

Summary

In today’s blog post I demonstrated how to install the Keras deep learning library using the TensorFlow backend. Previous blog posts have discussed how to install Keras with a Theano backend.

When it comes to choosing a backend for Keras you need to consider a few aspects.

The first is the popularity and therefore the probability that a given library will continue to be updated and supported in the future. In this case, TensorFlow wins hands down — it is currently the most popular numerical computation engine in the world used for machine learning and deep learning.

Secondly, you need to consider the functionality of a given library. While Theano is just as easy to use as TensorFlow out-of-the-box (in terms of Keras backends), TensorFlow allows for a more architecture agnostic deployment. By using TensorFlow it becomes possible to train distributed deep learning networks across CPUs, GPUs, and other devices all without having to change a single line of code.

In my particular case I’m using both Theano and TensorFlow in my experiments. Future blog posts will provide updates with my TensorFlow experience.

In the meantime, I suggest that you install both Theano and TensorFlow on your system and determine for yourself which backend is most suitable for your needs.

If you enjoyed this install tutorial and found it helpful be sure to leave a note in the comments!

And if you would like to receive email updates when new blog posts are published on the PyImageSearch blog, please enter your email address in the form below.

The post Installing Keras with TensorFlow backend appeared first on PyImageSearch.

from PyImageSearch http://ift.tt/2fySjh6
via IFTTT