videoclassification

Use Tensforflow frozen graph for video classification

TensorFlow C++ and Python Video Classification Demo

This example shows how you can load a pre-trained TensorFlow network and use it to recognize objects in images/videos in Python/C++.

##TODO: Make changes to C++ to support video classification

Description

This demo uses a Google Inception model to classify image/video files that are passed in on the command line.

Clone this repo

$ git clone https://github.com/M-Melodious/videoclassification.git
$ cd videoclassification

To build/install/run

The TensorFlow GraphDef that contains the model definition and weights is not packaged in the repo because of its size. Instead, you must first download the file to the data directory in the source tree:

$ curl -L "https://storage.googleapis.com/download.tensorflow.org/models/inception_v3_2016_08_28_frozen.pb.tar.gz" |
  tar -C videoclassification/label_image/data -xz

Then, as long as you've managed to build the main TensorFlow framework, you should have everything you need to run this example installed already.

Once extracted, see the labels file in the data directory for the possible classifications, which are the 1,000 categories used in the Imagenet competition.

To build it, run this command:

$ bazel build ...

That should build a binary executable that you can then run like this:

$ bazel-bin/label_image/label_image

##NOTE: Tensorflow and C++ will not run, as I've removed it from WORKSPACE

This uses the default example image that ships with the framework, and should output something similar to this:

I tensorflow/examples/label_image/main.cc:206] military uniform (653): 0.834306
I tensorflow/examples/label_image/main.cc:206] mortarboard (668): 0.0218692
I tensorflow/examples/label_image/main.cc:206] academic gown (401): 0.0103579
I tensorflow/examples/label_image/main.cc:206] pickelhaube (716): 0.00800814
I tensorflow/examples/label_image/main.cc:206] bulletproof vest (466): 0.00535088

In this case, we're using the default image of Admiral Grace Hopper, and you can see the network correctly spots she's wearing a military uniform, with a high score of 0.8.

Next, try it out on your own images by supplying the --image= argument, e.g.

$ bazel-bin/label_image/label_image --image=my_image.png

For a more detailed look at this code, you can check out the C++ section of the Inception tutorial.

Python implementation

label_image.py is a python implementation that provides code corresponding to the C++ code here. This gives more intuitive mapping between C++ and Python than the Python code mentioned in the Inception tutorial. and could be easier to add visualization or debug code.

bazel-bin/label_image/label_image_py should be there after

$ bazel build ...

You can provide either image or video. If neither is provided, it will use default image.

Run

$ bazel-bin/label_image/label_image_py -v path/to/video.mp4
$ bazel-bin/label_image/label_image_py -i path/to/image.jpg

And get result similar to this:

Class: military uniform Score: 83.4305%