Keras VGGFace extracting features

0
votes

I am trying to extract features from a convolution layer of the VGGFace model, using TensorFlow & Keras.

This is my code:

# Layer Features
layer_name = 'conv1_2' # Edit this line
vgg_model = VGGFace() # Pooling: None, avg or max
out = vgg_model.get_layer(layer_name).output
vgg_model_new = Model(vgg_model.input, out)

def main():
    img = image.load_img('myimage.jpg', target_size=(224, 224))
    x = image.img_to_array(img)
    x = np.expand_dims(x, axis=0)
    x = utils.preprocess_input(x, version=1)
    preds = vgg_model_new.predict(x)
    print('Predicted:', utils.decode_predictions(preds))
    exit(0)

However, at the print('Predicted:', utils.decode_predictions(preds)) line I am getting the following error:

Message=decode_predictions expects a batch of predictions (i.e. a 2D array of shape (samples, 2622)) for V1 or (samples, 8631) for V2.Found array with shape: (1, 224, 224, 64)

I just want to extract features, I don't need to classify my images at this point. This code is based on https://github.com/rcmalli/keras-vggface

1
pythontensorflowkerasvgg-net

1 Answers

1
votes

You shouldn't use utils.decode_predictions(preds) there because it's only for classification. You can see the definition of the function here https://github.com/rcmalli/keras-vggface/blob/master/keras_vggface/utils.py#L66

If you want to print the features, use print('Predicted:',preds)