All geek questions in one place

Adding a resize layer to a keras sequential model

How to add a layer with dimensions in

model = Sequential()

using

model.add(...)

To resize an image from a form (160, 320, 3) - (224,224.3)?

+8
source share
4 answers

Usually use a layer for this Reshape:

model.add(Reshape((224,224,3), input_shape=(160,320,3))

but since your target sizes do not allow you to store all the data from the input measurements ( 224*224 != 160*320), this will not work. You can use only Reshapeif the number of elements does not change.

If everything is fine with the loss of some data in your image, you can specify your own version with losses:

model.add(Reshape(-1,3), input_shape=(160,320,3))
model.add(Lambda(lambda x: x[:50176])) # throw away some, so that #data = 224^2
model.add(Reshape(224,224,3))

, , , .

+2

, resize_images tensorflow.

https://www.tensorflow.org/api_docs/python/tf/image/resize_images

, keras , , , , theano. keras, . , .

import keras
import keras.backend as K
from keras.utils import conv_utils
from keras.engine import InputSpec
from keras.engine import Layer
from tensorflow import image as tfi

class ResizeImages(Layer):
    """Resize Images to a specified size

    # Arguments
        output_size: Size of output layer width and height
        data_format: A string,
            one of 'channels_last' (default) or 'channels_first'.
            The ordering of the dimensions in the inputs.
            'channels_last' corresponds to inputs with shape
            '(batch, height, width, channels)' while 'channels_first'
            corresponds to inputs with shape
            '(batch, channels, height, width)'.
            It defaults to the 'image_data_format' value found in your
            Keras config file at '~/.keras/keras.json'.
            If you never set it, then it will be "channels_last".

    # Input shape
        - If 'data_format='channels_last'':
            4D tensor with shape:
            '(batch_size, rows, cols, channels)'
        - If 'data_format='channels_first'':
            4D tensor with shape:
            '(batch_size, channels, rows, cols)'

    # Output shape
        - If 'data_format='channels_last'':
            4D tensor with shape:
            '(batch_size, pooled_rows, pooled_cols, channels)'
        - If 'data_format='channels_first'':
            4D tensor with shape:
            '(batch_size, channels, pooled_rows, pooled_cols)'
    """
    def __init__(self, output_dim=(1, 1), data_format=None, **kwargs):
        super(ResizeImages, self).__init__(**kwargs)
        data_format = conv_utils.normalize_data_format(data_format)
        self.output_dim = conv_utils.normalize_tuple(output_dim, 2, 'output_dim')
        self.data_format = conv_utils.normalize_data_format(data_format)
        self.input_spec = InputSpec(ndim=4)

    def build(self, input_shape):
        self.input_spec = [InputSpec(shape=input_shape)]

    def compute_output_shape(self, input_shape):
        if self.data_format == 'channels_first':
            return (input_shape[0], input_shape[1], self.output_dim[0], self.output_dim[1])
        elif self.data_format == 'channels_last':
            return (input_shape[0], self.output_dim[0], self.output_dim[1], input_shape[3])

    def _resize_fun(self, inputs, data_format):
        try:
            assert keras.backend.backend() == 'tensorflow'
            assert self.data_format == 'channels_last'
        except AssertionError:
            print "Only tensorflow backend is supported for the resize layer and accordingly 'channels_last' ordering"
        output = tfi.resize_images(inputs, self.output_dim)
        return output

    def call(self, inputs):
        output = self._resize_fun(inputs=inputs, data_format=self.data_format)
        return output

    def get_config(self):
        config = {'output_dim': self.output_dim,
                  'padding': self.padding,
                  'data_format': self.data_format}
        base_config = super(ResizeImages, self).get_config()
        return dict(list(base_config.items()) + list(config.items()))
+9

Reshape, NumPy reshape, 4x4 2x8, :

0 0 0 0
1 1 1 1    ->    0 0 0 0 1 1 1 1
2 2 2 2          2 2 2 2 3 3 3 3
3 3 3 3

/"" , , Tensorflows image_resize. ! , -:

model.add( keras.layers.Lambda( 
    lambda image: tf.image.resize_images( 
        image, 
        (224, 224), 
        method = tf.image.ResizeMethod.BICUBIC,
        align_corners = True, # possibly important
        preserve_aspect_ratio = True
    )
))

, 160x320, , . , , .

+2

@KeithWM output_scale, output_scale = 2, , 2 :)

class ResizeImages(Layer):
    """Resize Images to a specified size
    https://stackoverflow.com/questions/41903928/add-a-resizing-layer-to-a-keras-sequential-model

    # Arguments
        output_dim: Size of output layer width and height
        output_scale: scale compared with input
        data_format: A string,
            one of 'channels_last' (default) or 'channels_first'.
            The ordering of the dimensions in the inputs.
            'channels_last' corresponds to inputs with shape
            '(batch, height, width, channels)' while 'channels_first'
            corresponds to inputs with shape
            '(batch, channels, height, width)'.
            It defaults to the 'image_data_format' value found in your
            Keras config file at '~/.keras/keras.json'.
            If you never set it, then it will be "channels_last".

    # Input shape
        - If 'data_format='channels_last'':
            4D tensor with shape:
            '(batch_size, rows, cols, channels)'
        - If 'data_format='channels_first'':
            4D tensor with shape:
            '(batch_size, channels, rows, cols)'

    # Output shape
        - If 'data_format='channels_last'':
            4D tensor with shape:
            '(batch_size, pooled_rows, pooled_cols, channels)'
        - If 'data_format='channels_first'':
            4D tensor with shape:
            '(batch_size, channels, pooled_rows, pooled_cols)'
    """

    def __init__(self, output_dim=(1, 1), output_scale=None, data_format=None, **kwargs):
        super(ResizeImages, self).__init__(**kwargs)
        data_format = normalize_data_format(data_format)  # does not have
        self.naive_output_dim = conv_utils.normalize_tuple(output_dim,
                                                           2, 'output_dim')
        self.naive_output_scale = output_scale
        self.data_format = normalize_data_format(data_format)
        self.input_spec = InputSpec(ndim=4)

    def build(self, input_shape):
        self.input_spec = [InputSpec(shape=input_shape)]
        if self.naive_output_scale is not None:
            if self.data_format == 'channels_first':
                self.output_dim = (self.naive_output_scale * input_shape[2],
                                   self.naive_output_scale * input_shape[3])
            elif self.data_format == 'channels_last':
                self.output_dim = (self.naive_output_scale * input_shape[1],
                                   self.naive_output_scale * input_shape[2])
        else:
            self.output_dim = self.naive_output_dim

    def compute_output_shape(self, input_shape):
        if self.data_format == 'channels_first':
            return (input_shape[0], input_shape[1], self.output_dim[0], self.output_dim[1])
        elif self.data_format == 'channels_last':
            return (input_shape[0], self.output_dim[0], self.output_dim[1], input_shape[3])

    def _resize_fun(self, inputs, data_format):
        try:
            assert keras.backend.backend() == 'tensorflow'
            assert self.data_format == 'channels_last'
        except AssertionError:
            print("Only tensorflow backend is supported for the resize layer and accordingly 'channels_last' ordering")
        output = tf.image.resize_images(inputs, self.output_dim)
        return output

    def call(self, inputs):
        output = self._resize_fun(inputs=inputs, data_format=self.data_format)
        return output

    def get_config(self):
        config = {'output_dim': self.output_dim,
                  'padding': self.padding,
                  'data_format': self.data_format}
        base_config = super(ResizeImages, self).get_config()
        return dict(list(base_config.items()) + list(config.items()))
0

Source: https://habr.com/ru/post/1668055/

More articles:


All Articles