Get class labels from Keras functional model

Question

I have a functional model in Keras  Resnet50 from repo examples   I trained it with ImageDataGenerator and flow from directory data and saved model to  h5 file  When I call model predict I get an array of class probabilities  But I want to associate them with class labels  in my case - folder names   How can I get them  I found that I could use model predict classes and model predict proba  but I don t have these functions in Functional model  only in Sequential

User · Answer

To map predicted classes and filenames using ImageDataGenerator, I use:

# Data generator and prediction
test_datagen = ImageDataGenerator(rescale=1./255)
test_generator = test_datagen.flow_from_directory(
        inputpath,
        target_size=(150, 150),
        batch_size=20,
        class_mode='categorical',
        shuffle=False)
pred = model.predict_generator(test_generator, steps=len(test_generator), verbose=0)
# Get classes by max element in np (as a list)
classes = list(np.argmax(pred, axis=1))
# Get filenames (set shuffle=false in generator is important)
filenames = test_generator.filenames

I can loop over predicted classes and the associated filename using:

for f in zip(classes, filenames):
    ...

User · Answer

UPDATE  This is no longer valid for newer Keras versions  Please use argmax   as in the answer from Emilia Apostolova   The functional API models have just the predict   function which for classification would return the class probabilities  You can then select the most probable classes using the probas to classes   utility function  Example   y proba   model predict x  y classes   keras np utils probas to classes y proba    This is equivalent to model predict classes x  on the Sequential model   The reason for this is that the functional API support more general class of tasks where predict classes   would not make sense   More info  https   github com fchollet keras issues 2524

User · Answer

In addition to  Emilia Apostolova answer to get the ground truth labels  from    generator   train datagen flow from directory  train   batch size batch size    just call  y true labels   generator classes

User · Answer

When one uses flow from directory the problem is how to interpret the probability outputs  As in  how to map the probability outputs and the class labels as how flow from directory creates one-hot vectors is not known in prior   We can get a dictionary that maps the class labels to the index of the prediction vector that we get as the output when we use   generator  train datagen flow from directory  train   batch size batch size  label map    generator class indices    The label map variable is a dictionary like this    class 14   5   class 10   1   class 11   2   class 12   3   class 13   4   class 2   6   class 3   7   class 1   0   class 6   10   class 7   11   class 4   8   class 5   9   class 8   12   class 9   13    Then from this the relation can be derived between the probability scores and class names   Basically  you can create this dictionary by this code   from glob import glob class names   glob        Reads all the folders in which images are present class names   sorted class names    Sorting them name id map   dict zip class names  range len class names       The variable name id map in the above code also contains the same dictionary as the one obtained from class indices function of flow from directory   Hope this helps

User · Answer

It is possible to save a  list  of labels in keras model directly  This way the user who uses the model for predictions and does not have any other sources of information can perform the lookup himself  Here is a dummy example of how one can perform an  injection  of labels    assume we get labels as list labels     cat   dog   horse   tomato     here we start building our model with input image 299x299 and one output layer xx   Input shape  299 299 3   flat   Flatten   xx  output   Dense shape  4   flat    here we perform injection of labels tf labels   tf constant  labels  dtype  string   tf labels   tf tile labels  tf shape xx  0  1   output labels   Lambda lambda x  tf labels name  label injection   xx   and finaly creating a model model tf keras Model xx  output output labels     When used for prediction  this model returns tensor of scores and tensot of string labels  Model like this can be saved to h5  In this case the file contains the labels  This model can also be exported to saved model and used for serving in the cloud

User · Answer

You must use the labels index you have  here what I do for text classification     data labels    1  2  1     labels index      website    0   money    1          to feed model label categories   to categorical np asarray labels      Then  for predictions   texts     hello  rejoins moi sur skype    bonjour comment   a va      tu me donnes de l argent    sequences   tokenizer texts to sequences texts   data   pad sequences sequences  maxlen MAX SEQUENCE LENGTH   predictions   model predict data   t   0  for text in texts      i   0     print  Prediction for    s         text       for label in labels index          print   t s    gt   f     label  predictions t  i            i   i   1     t   t   1   This gives   Prediction for  hello  rejoins moi sur skype        website    gt  0 759483     money    gt  0 037091     under    gt  0 010587     camsite    gt  0 114436     email    gt  0 075975     abuse    gt  0 002428 Prediction for  bonjour comment   a va          website    gt  0 433079     money    gt  0 084878     under    gt  0 048375     camsite    gt  0 036674     email    gt  0 369197     abuse    gt  0 027798 Prediction for  tu me donnes de l argent        website    gt  0 006223     money    gt  0 095308     under    gt  0 003586     camsite    gt  0 003115     email    gt  0 884112     abuse    gt  0 007655

User · Answer

y prob   model predict x   y classes   y prob argmax axis -1    As suggested here

[python] Get class labels from Keras functional model

Examples related to python

Examples related to keras