Transformers documentation

Utilities for Image Processors

Join the Model Database community

and get access to the augmented documentation experience

Collaborate on models, datasets and Spaces

Faster examples with accelerated inference

Switch between documentation themes

to get started

Utilities for Image Processors

This page lists all the utility functions used by the image processors, mainly the functional transformations used to process the images.

Most of those are only useful if you are studying the code of the image processors in the library.

Image Transformations

transformers.image_transforms.center_crop

< source >

( image: ndarray size: typing.Tuple[int, int] data_format: typing.Union[transformers.image_utils.ChannelDimension, str, NoneType] = None input_data_format: typing.Union[transformers.image_utils.ChannelDimension, str, NoneType] = None return_numpy: typing.Optional[bool] = None ) → np.ndarray

Parameters

image (np.ndarray) — The image to crop.
size (Tuple[int, int]) — The target size for the cropped image.
data_format (str or ChannelDimension, optional) — The channel dimension format for the output image. Can be one of:
- "channels_first" or ChannelDimension.FIRST: image in (num_channels, height, width) format.
- "channels_last" or ChannelDimension.LAST: image in (height, width, num_channels) format. If unset, will use the inferred format of the input image.
input_data_format (str or ChannelDimension, optional) — The channel dimension format for the input image. Can be one of:
- "channels_first" or ChannelDimension.FIRST: image in (num_channels, height, width) format.
- "channels_last" or ChannelDimension.LAST: image in (height, width, num_channels) format. If unset, will use the inferred format of the input image.
return_numpy (bool, optional) — Whether or not to return the cropped image as a numpy array. Used for backwards compatibility with the previous ImageFeatureExtractionMixin method.
- Unset: will return the same type as the input image.
- True: will return a numpy array.
- False: will return a PIL.Image.Image object.

Returns

np.ndarray

The cropped image.

Crops the image to the specified size using a center crop. Note that if the image is too small to be cropped to the size given, it will be padded (so the returned result will always be of size size).

transformers.image_transforms.center_to_corners_format

< source >

( bboxes_center: TensorType )

Converts bounding boxes from center format to corners format.

center format: contains the coordinate for the center of the box and its width, height dimensions (center_x, center_y, width, height) corners format: contains the coodinates for the top-left and bottom-right corners of the box (top_left_x, top_left_y, bottom_right_x, bottom_right_y)

transformers.image_transforms.corners_to_center_format

< source >

( bboxes_corners: TensorType )

Converts bounding boxes from corners format to center format.

corners format: contains the coodinates for the top-left and bottom-right corners of the box (top_left_x, top_left_y, bottom_right_x, bottom_right_y) center format: contains the coordinate for the center of the box and its the width, height dimensions (center_x, center_y, width, height)

transformers.image_transforms.id_to_rgb

< source >

( id_map )

Converts unique ID to RGB color.

transformers.image_transforms.normalize

< source >

( image: ndarray mean: typing.Union[float, typing.Iterable[float]] std: typing.Union[float, typing.Iterable[float]] data_format: typing.Optional[transformers.image_utils.ChannelDimension] = None input_data_format: typing.Union[transformers.image_utils.ChannelDimension, str, NoneType] = None )

Parameters

image (np.ndarray) — The image to normalize.
mean (float or Iterable[float]) — The mean to use for normalization.
std (float or Iterable[float]) — The standard deviation to use for normalization.
data_format (ChannelDimension, optional) — The channel dimension format of the output image. If unset, will use the inferred format from the input.
input_data_format (ChannelDimension, optional) — The channel dimension format of the input image. If unset, will use the inferred format from the input.

Normalizes image using the mean and standard deviation specified by mean and std.

image = (image - mean) / std

transformers.image_transforms.pad

< source >

( image: ndarray padding: typing.Union[int, typing.Tuple[int, int], typing.Iterable[typing.Tuple[int, int]]] mode: PaddingMode = <PaddingMode.CONSTANT: 'constant'> constant_values: typing.Union[float, typing.Iterable[float]] = 0.0 data_format: typing.Union[transformers.image_utils.ChannelDimension, str, NoneType] = None input_data_format: typing.Union[transformers.image_utils.ChannelDimension, str, NoneType] = None ) → np.ndarray

Parameters

image (np.ndarray) — The image to pad.
padding (int or Tuple[int, int] or Iterable[Tuple[int, int]]) — Padding to apply to the edges of the height, width axes. Can be one of three formats:
- ((before_height, after_height), (before_width, after_width)) unique pad widths for each axis.
- ((before, after),) yields same before and after pad for height and width.
- (pad,) or int is a shortcut for before = after = pad width for all axes.
mode (PaddingMode) — The padding mode to use. Can be one of:
- "constant": pads with a constant value.
- "reflect": pads with the reflection of the vector mirrored on the first and last values of the vector along each axis.
- "replicate": pads with the replication of the last value on the edge of the array along each axis.
- "symmetric": pads with the reflection of the vector mirrored along the edge of the array.
constant_values (float or Iterable[float], optional) — The value to use for the padding if mode is "constant".
data_format (str or ChannelDimension, optional) — The channel dimension format for the output image. Can be one of:
- "channels_first" or ChannelDimension.FIRST: image in (num_channels, height, width) format.
- "channels_last" or ChannelDimension.LAST: image in (height, width, num_channels) format. If unset, will use same as the input image.
input_data_format (str or ChannelDimension, optional) — The channel dimension format for the input image. Can be one of:
- "channels_first" or ChannelDimension.FIRST: image in (num_channels, height, width) format.
- "channels_last" or ChannelDimension.LAST: image in (height, width, num_channels) format. If unset, will use the inferred format of the input image.

Returns

np.ndarray

The padded image.

Pads the image with the specified (height, width) padding and mode.

transformers.image_transforms.rgb_to_id

< source >

( color )

Converts RGB color to unique ID.

transformers.image_transforms.rescale

< source >

( image: ndarray scale: float data_format: typing.Optional[transformers.image_utils.ChannelDimension] = None dtype: dtype = <class 'numpy.float32'> input_data_format: typing.Union[transformers.image_utils.ChannelDimension, str, NoneType] = None ) → np.ndarray

Parameters

image (np.ndarray) — The image to rescale.
scale (float) — The scale to use for rescaling the image.
data_format (ChannelDimension, optional) — The channel dimension format of the image. If not provided, it will be the same as the input image.
dtype (np.dtype, optional, defaults to np.float32) — The dtype of the output image. Defaults to np.float32. Used for backwards compatibility with feature extractors.
input_data_format (ChannelDimension, optional) — The channel dimension format of the input image. If not provided, it will be inferred from the input image.

Returns

np.ndarray

The rescaled image.

Rescales image by scale.

transformers.image_transforms.resize

< source >

( image size: typing.Tuple[int, int] resample: PILImageResampling = None reducing_gap: typing.Optional[int] = None data_format: typing.Optional[transformers.image_utils.ChannelDimension] = None return_numpy: bool = True input_data_format: typing.Union[transformers.image_utils.ChannelDimension, str, NoneType] = None ) → np.ndarray

Parameters

image (PIL.Image.Image or np.ndarray or torch.Tensor) — The image to resize.
size (Tuple[int, int]) — The size to use for resizing the image.
resample (int, optional, defaults to PILImageResampling.BILINEAR) — The filter to user for resampling.
reducing_gap (int, optional) — Apply optimization by resizing the image in two steps. The bigger reducing_gap, the closer the result to the fair resampling. See corresponding Pillow documentation for more details.
data_format (ChannelDimension, optional) — The channel dimension format of the output image. If unset, will use the inferred format from the input.
return_numpy (bool, optional, defaults to True) — Whether or not to return the resized image as a numpy array. If False a PIL.Image.Image object is returned.
input_data_format (ChannelDimension, optional) — The channel dimension format of the input image. If unset, will use the inferred format from the input.

Returns

np.ndarray

The resized image.

Resizes image to (height, width) specified by size using the PIL library.

transformers.image_transforms.to_pil_image

< source >

( image: typing.Union[numpy.ndarray, ForwardRef('PIL.Image.Image'), ForwardRef('torch.Tensor'), ForwardRef('tf.Tensor'), ForwardRef('jnp.ndarray')] do_rescale: typing.Optional[bool] = None input_data_format: typing.Union[transformers.image_utils.ChannelDimension, str, NoneType] = None ) → PIL.Image.Image

Parameters

image (PIL.Image.Image or numpy.ndarray or torch.Tensor or tf.Tensor) — The image to convert to the PIL.Image format.
do_rescale (bool, optional) — Whether or not to apply the scaling factor (to make pixel values integers between 0 and 255). Will default to True if the image type is a floating type and casting to int would result in a loss of precision, and False otherwise.
input_data_format (ChannelDimension, optional) — The channel dimension format of the input image. If unset, will use the inferred format from the input.

Returns

PIL.Image.Image

The converted image.

Converts image to a PIL Image. Optionally rescales it and puts the channel dimension back as the last axis if needed.

ImageProcessingMixin

class transformers.ImageProcessingMixin

< source >

( **kwargs )

This is an image processor mixin used to provide saving/loading functionality for sequential and image feature extractors.

fetch_images

< source >

( image_url_or_urls: typing.Union[str, typing.List[str]] )

Convert a single or a list of urls into the corresponding PIL.Image objects.

If a single url is passed, the return value will be a single object. If a list is passed a list of objects is returned.

from_dict

< source >

( image_processor_dict: typing.Dict[str, typing.Any] **kwargs ) → ImageProcessingMixin

Parameters

image_processor_dict (Dict[str, Any]) — Dictionary that will be used to instantiate the image processor object. Such a dictionary can be retrieved from a pretrained checkpoint by leveraging the to_dict() method.
kwargs (Dict[str, Any]) — Additional parameters from which to initialize the image processor object.

Returns

ImageProcessingMixin

The image processor object instantiated from those parameters.

Instantiates a type of ImageProcessingMixin from a Python dictionary of parameters.

from_json_file

< source >

( json_file: typing.Union[str, os.PathLike] ) → A image processor of type ImageProcessingMixin

Parameters

json_file (str or os.PathLike) — Path to the JSON file containing the parameters.

Returns

A image processor of type ImageProcessingMixin

The image_processor object instantiated from that JSON file.

Instantiates a image processor of type ImageProcessingMixin from the path to a JSON file of parameters.

from_pretrained

< source >

( pretrained_model_name_or_path: typing.Union[str, os.PathLike] cache_dir: typing.Union[str, os.PathLike, NoneType] = None force_download: bool = False local_files_only: bool = False token: typing.Union[str, bool, NoneType] = None revision: str = 'main' **kwargs )

Parameters

pretrained_model_name_or_path (str or os.PathLike) — This can be either:
- a string, the model id of a pretrained image_processor hosted inside a model repo on huggingface.co. Valid model ids can be located at the root-level, like bert-base-uncased, or namespaced under a user or organization name, like dbmdz/bert-base-german-cased.
- a path to a directory containing a image processor file saved using the save_pretrained() method, e.g., ./my_model_directory/.
- a path or url to a saved image processor JSON file, e.g., ./my_model_directory/preprocessor_config.json.
cache_dir (str or os.PathLike, optional) — Path to a directory in which a downloaded pretrained model image processor should be cached if the standard cache should not be used.
force_download (bool, optional, defaults to False) — Whether or not to force to (re-)download the image processor files and override the cached versions if they exist.
resume_download (bool, optional, defaults to False) — Whether or not to delete incompletely received file. Attempts to resume the download if such a file exists.
proxies (Dict[str, str], optional) — A dictionary of proxy servers to use by protocol or endpoint, e.g., {'http': 'foo.bar:3128', 'http://hostname': 'foo.bar:4012'}. The proxies are used on each request.
token (str or bool, optional) — The token to use as HTTP bearer authorization for remote files. If True, or not specified, will use the token generated when running huggingface-cli login (stored in ~/.huggingface).
revision (str, optional, defaults to "main") — The specific model version to use. It can be a branch name, a tag name, or a commit id, since we use a git-based system for storing models and other artifacts on huggingface.co, so revision can be any identifier allowed by git.

Instantiate a type of ImageProcessingMixin from an image processor.

Examples:

# We can't instantiate directly the base class *ImageProcessingMixin* so let's show the examples on a
# derived class: *CLIPImageProcessor*
image_processor = CLIPImageProcessor.from_pretrained(
    "openai/clip-vit-base-patch32"
)  # Download image_processing_config from huggingface.co and cache.
image_processor = CLIPImageProcessor.from_pretrained(
    "./test/saved_model/"
)  # E.g. image processor (or model) was saved using *save_pretrained('./test/saved_model/')*
image_processor = CLIPImageProcessor.from_pretrained("./test/saved_model/preprocessor_config.json")
image_processor = CLIPImageProcessor.from_pretrained(
    "openai/clip-vit-base-patch32", do_normalize=False, foo=False
)
assert image_processor.do_normalize is False
image_processor, unused_kwargs = CLIPImageProcessor.from_pretrained(
    "openai/clip-vit-base-patch32", do_normalize=False, foo=False, return_unused_kwargs=True
)
assert image_processor.do_normalize is False
assert unused_kwargs == {"foo": False}

get_image_processor_dict

< source >

( pretrained_model_name_or_path: typing.Union[str, os.PathLike] **kwargs ) → Tuple[Dict, Dict]

Parameters

pretrained_model_name_or_path (str or os.PathLike) — The identifier of the pre-trained checkpoint from which we want the dictionary of parameters.
subfolder (str, optional, defaults to "") — In case the relevant files are located inside a subfolder of the model repo on huggingface.co, you can specify the folder name here.

Returns

Tuple[Dict, Dict]

The dictionary(ies) that will be used to instantiate the image processor object.

From a pretrained_model_name_or_path, resolve to a dictionary of parameters, to be used for instantiating a image processor of type ~image_processor_utils.ImageProcessingMixin using from_dict.

push_to_hub

< source >

( repo_id: str use_temp_dir: typing.Optional[bool] = None commit_message: typing.Optional[str] = None private: typing.Optional[bool] = None token: typing.Union[bool, str, NoneType] = None max_shard_size: typing.Union[int, str, NoneType] = '10GB' create_pr: bool = False safe_serialization: bool = False revision: str = None **deprecated_kwargs )

Parameters

repo_id (str) — The name of the repository you want to push your image processor to. It should contain your organization name when pushing to a given organization.
use_temp_dir (bool, optional) — Whether or not to use a temporary directory to store the files saved before they are pushed to the Hub. Will default to True if there is no directory named like repo_id, False otherwise.
commit_message (str, optional) — Message to commit while pushing. Will default to "Upload image processor".
private (bool, optional) — Whether or not the repository created should be private.
token (bool or str, optional) — The token to use as HTTP bearer authorization for remote files. If True, will use the token generated when running huggingface-cli login (stored in ~/.huggingface). Will default to True if repo_url is not specified.
max_shard_size (int or str, optional, defaults to "10GB") — Only applicable for models. The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size lower than this size. If expressed as a string, needs to be digits followed by a unit (like "5MB").
create_pr (bool, optional, defaults to False) — Whether or not to create a PR with the uploaded files or directly commit.
safe_serialization (bool, optional, defaults to False) — Whether or not to convert the model weights in safetensors format for safer serialization.
revision (str, optional) — Branch to push the uploaded files to.

Upload the image processor file to the 🤗 Model Hub.

Examples:

from transformers import AutoImageProcessor

image processor = AutoImageProcessor.from_pretrained("bert-base-cased")

# Push the image processor to your namespace with the name "my-finetuned-bert".
image processor.push_to_hub("my-finetuned-bert")

# Push the image processor to an organization with the name "my-finetuned-bert".
image processor.push_to_hub("huggingface/my-finetuned-bert")

register_for_auto_class

< source >

( auto_class = 'AutoImageProcessor' )

Parameters

auto_class (str or type, optional, defaults to "AutoImageProcessor ") — The auto class to register this new image processor with.

Register this class with a given auto class. This should only be used for custom image processors as the ones in the library are already mapped with AutoImageProcessor .

This API is experimental and may have some slight breaking changes in the next releases.

save_pretrained

< source >

( save_directory: typing.Union[str, os.PathLike] push_to_hub: bool = False **kwargs )

Parameters

save_directory (str or os.PathLike) — Directory where the image processor JSON file will be saved (will be created if it does not exist).
push_to_hub (bool, optional, defaults to False) — Whether or not to push your model to the Model Database model hub after saving it. You can specify the repository you want to push to with repo_id (will default to the name of save_directory in your namespace).
kwargs (Dict[str, Any], optional) — Additional key word arguments passed along to the push_to_hub() method.