Prefix to use for filenames of saved pictures (only relevant if `save_to_dir` is set). This allows you to optionally specify a directory to which to save the augmented pictures being generated. save_to_dir: None or str (default: None).seed: Optional random seed for shuffling and transformations.If set to False, sorts the data in alphanumeric order. shuffle: Whether to shuffle the data (default: True).batch_size: Size of the batches of data (default: 32).
KERAS IMAGE DATA GENERATOR GENERATOR
If None, no labels are returned(the generator will only yield batches of image data, which is useful to use with `model.predict_generator()`). Determines the type of label arrays that are returned: - "categorical" will be 2D one-hot encoded labels, - "binary" will be 1D binary labels, - "sparse" will be 1D integer labels, - "input" will be images identical to input images (mainly used to work with autoencoders). class_mode: One of "categorical", "binary", "sparse", "input", or None.The dictionary containing the mapping from class names to class indices can be obtained via the attribute `class_indices`. If not provided, the list of classes will be automatically inferred from the subdirectory names/structure under `directory`, where each subdirectory will be treated as a different class(and the order of the classes, which will map to the label indices, will be alphanumeric). classes: Optional list of class subdirectories(e.g.color_mode: One of "grayscale", "rgb", "rgba".The dimensions to which all images found will be resized. target_size: Tuple of integers `(height, width)`, default: `(256, 256)`.Any PNG, JPG, BMP, PPM or TIF imagesinside each of the subdirectories directory tree will be included in the generator. It should contain one subdirectory per class. directory: string, path to the target directory.Takes the path to a directory & generates batches of augmented data. dtype: Dtype to use for the generated arrays.Fraction of images reserved for validation (strictly between 0 and 1). If you never set it, then it will be "channels_last". It defaults to the image_data_format value found in your Keras config file at ~/.keras/keras.json. "channels_last" mode means that the images should have shape (samples, height, width, channels), "channels_first" mode means that the images should have shape (samples, channels, height, width). data_format: Image data format, either "channels_first" or "channels_last".The function should take one argument: one image (Numpy tensor with rank 3), and should output a Numpy tensor with the same shape. The function will run after the image is resized and augmented. preprocessing_function: function that will be implied on each input.If None or 0, no rescaling is applied, otherwise we multiply the data by the value provided (after applying all other transformations). Value used for points outside the boundaries when fill_mode = "constant". 'constant': kkkkkkkk|abcd|kkkkkkkk (cval=k).Points outside the boundaries of the input are filled according to the given mode: Shear Intensity (Shear angle in counter-clockwise direction in degrees) Range for picking a brightness shift value from. brightness_range: Tuple or list of two floats.height_shift_range: Float, 1-D array-like or int.With width_shift_range=2 possible values are integers, same as with width_shift_range=, while with width_shift_range=1.0 possible values are floats in the interval [-1.0, +1.0).int: integer number of pixels from interval (-width_shift_range, +width_shift_range).1-D array-like: random elements from the array.float: fraction of total width, if = 1.width_shift_range: Float, 1-D array-like or int.zca_epsilon: epsilon for ZCA whitening.samplewise_std_normalization: Boolean.Divide inputs by std of the dataset, feature-wise.0 featurewise_std_normalization: Boolean.Set input mean to 0 over the dataset, feature-wise.
0 kommentar(er)
