fiftyone.utils.data.parsers

Sample parsers.

Copyright 2017-2025, Voxel51, Inc.

voxel51.com

Functions:

add_images(dataset, samples, sample_parser)	Adds the given images to the dataset.
add_labeled_images(dataset, samples, ...[, ...])	Adds the given labeled images to the dataset.
add_videos(dataset, samples, sample_parser)	Adds the given videos to the dataset.
add_labeled_videos(dataset, samples, ...[, ...])	Adds the given labeled videos to the dataset.

Classes:

SampleParser()	Base interface for sample parsers.
UnlabeledImageSampleParser()	Interface for SampleParser instances that parse unlabeled image samples.
UnlabeledVideoSampleParser()	Interface for SampleParser instances that parse unlabeled video samples.
UnlabeledMediaSampleParser()	Interface for SampleParser instances that parse unlabeled media samples.
ImageSampleParser()	Sample parser that parses unlabeled image samples.
VideoSampleParser()	Sample parser that parses unlabeled video samples.
MediaSampleParser()	Sample parser that parses unlabeled media samples.
LabeledImageSampleParser()	Interface for SampleParser instances that parse labeled image samples.
LabeledVideoSampleParser()	Interface for SampleParser instances that parse labeled video samples.
LabeledImageTupleSampleParser()	Generic sample parser that parses samples that are (image_or_path, label) tuples, where:
ImageClassificationSampleParser([classes])	Generic parser for image classification(s) samples whose labels are represented as fiftyone.core.labels.Classification instances.
ImageDetectionSampleParser([label_field, ...])	Generic parser for image detection samples whose labels are represented as fiftyone.core.labels.Detections instances.
ImageLabelsSampleParser([prefix, ...])	Generic parser for multitask image prediction samples whose labels are stored in eta.core.image.ImageLabels format.
FiftyOneImageClassificationSampleParser([...])	Parser for samples in FiftyOne image classification datasets.
FiftyOneTemporalDetectionSampleParser([...])	Parser for samples in FiftyOne temporal detection datasets.
FiftyOneImageDetectionSampleParser([classes])	Parser for samples in FiftyOne image detection datasets.
FiftyOneImageLabelsSampleParser([prefix, ...])	Parser for samples in FiftyOne image labels datasets.
VideoLabelsSampleParser([prefix, ...])	Generic parser for labeled video samples whose labels are represented in eta.core.video.VideoLabels format.
FiftyOneVideoLabelsSampleParser([prefix, ...])	Parser for samples in FiftyOne video labels datasets.
FiftyOneUnlabeledImageSampleParser([...])	Parser for fiftyone.core.sample.Sample instances that contain images.
FiftyOneLabeledImageSampleParser(label_field)	Parser for fiftyone.core.sample.Sample instances that contain labeled images.
ExtractClipsMixin([compute_metadata, ...])	Mixin for sample parsers that extract clips from fiftyone.core.clips.ClipView instances.
FiftyOneUnlabeledVideoSampleParser([...])	Parser for fiftyone.core.sample.Sample instances that contain videos.
FiftyOneLabeledVideoSampleParser([...])	Parser for fiftyone.core.sample.Sample instances that contain labeled videos.
FiftyOneUnlabeledMediaSampleParser([...])	Parser for fiftyone.core.sample.Sample instances that contain unlabeled media.

fiftyone.utils.data.parsers.add_images(dataset, samples, sample_parser, tags=None, generator=False, progress=None)

Adds the given images to the dataset.

This operation does not read the images.

See this guide for more details about adding images to a dataset by defining your own UnlabeledImageSampleParser.

Parameters:

• dataset – a fiftyone.core.dataset.Dataset

• samples – an iterable of samples that can be parsed by sample_parser

• sample_parser – a UnlabeledImageSampleParser instance to use to parse the samples

• tags (None) – an optional tag or iterable of tags to attach to each sample

• generator (False) – whether to yield ID batches as a generator as samples are added to the dataset

• progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns:

a list of IDs of the samples that were added to the dataset

fiftyone.utils.data.parsers.add_labeled_images(dataset, samples, sample_parser, label_field=None, tags=None, expand_schema=True, dynamic=False, generator=False, progress=None)

Adds the given labeled images to the dataset.

This operation will iterate over all provided samples, but the images will not be read (unless the sample parser requires it in order to compute image metadata).

See this guide for more details about adding labeled images to a dataset by defining your own LabeledImageSampleParser.

Parameters:

• dataset – a fiftyone.core.dataset.Dataset

• samples – an iterable of samples that can be parsed by sample_parser

• sample_parser – a LabeledImageSampleParser instance to use to parse the samples

• label_field (None) – controls the field(s) in which imported labels are stored. If the parser produces a single fiftyone.core.labels.Label instance per sample, this argument specifies the name of the field to use; the default is "ground_truth". If the parser produces a dictionary of labels per sample, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names

• tags (None) – an optional tag or iterable of tags to attach to each sample

• expand_schema (True) – whether to dynamically add new sample fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema

• dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered

• generator (False) – whether to yield ID batches as a generator as samples are added to the dataset

• progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns:

a list of IDs of the samples that were added to the dataset

fiftyone.utils.data.parsers.add_videos(dataset, samples, sample_parser, tags=None, generator=False, progress=None)

Adds the given videos to the dataset.

This operation does not read the videos.

See this guide for more details about adding videos to a dataset by defining your own UnlabeledVideoSampleParser.

Parameters:

• dataset – a fiftyone.core.dataset.Dataset

• samples – an iterable of samples that can be parsed by sample_parser

• sample_parser – a UnlabeledVideoSampleParser instance to use to parse the samples

• tags (None) – an optional tag or iterable of tags to attach to each sample

• generator (False) – whether to yield ID batches as a generator as samples are added to the dataset

• progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns:

a list of IDs of the samples that were added to the dataset

fiftyone.utils.data.parsers.add_labeled_videos(dataset, samples, sample_parser, label_field=None, tags=None, expand_schema=True, dynamic=False, generator=False, progress=None)

Adds the given labeled videos to the dataset.

This operation will iterate over all provided samples, but the videos will not be read/decoded/etc.

See this guide for more details about adding labeled videos to a dataset by defining your own LabeledVideoSampleParser.

Parameters:

• dataset – a fiftyone.core.dataset.Dataset

• samples – an iterable of samples that can be parsed by sample_parser

• sample_parser – a LabeledVideoSampleParser instance to use to parse the samples

• label_field (None) – controls the field(s) in which imported labels are stored. If the parser produces a single fiftyone.core.labels.Label instance per sample/frame, this argument specifies the name of the field to use; the default is "ground_truth". If the parser produces a dictionary of labels per sample/frame, this argument can be either a string prefix to prepend to each label key or a dict mapping label keys to field names; the default in this case is to directly use the keys of the imported label dictionaries as field names

• tags (None) – an optional tag or iterable of tags to attach to each sample

• expand_schema (True) – whether to dynamically add new sample fields encountered to the dataset schema. If False, an error is raised if a sample’s schema is not a subset of the dataset schema

• dynamic (False) – whether to declare dynamic attributes of embedded document fields that are encountered

• generator (False) – whether to yield ID batches as a generator as samples are added to the dataset

• progress (None) – whether to render a progress bar (True/False), use the default value fiftyone.config.show_progress_bars (None), or a progress callback function to invoke instead

Returns:

a list of IDs of the samples that were added to the dataset

class fiftyone.utils.data.parsers.SampleParser

Bases: object

Base interface for sample parsers.

SampleParser instances are used to parse samples emitted by dataset iterators when ingesting them into fiftyone.core.dataset.Dataset instances.

The general recipe for using SampleParser instances is as follows:

sample_parser = SampleParser(...)

for sample in samples:
    sample_parser.with_sample(sample)
    field = sample_parser.get_<field>()

where field is a subclass specific field to parse from the sample.

Attributes:

current_sample

The current sample.

Methods:

with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.
clear_sample()	Clears the current sample.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

class fiftyone.utils.data.parsers.UnlabeledImageSampleParser

Bases: SampleParser

Interface for SampleParser instances that parse unlabeled image samples.

Instances of this class must return images in numpy format.

The general recipe for using UnlabeledImageSampleParser instances is as follows:

sample_parser = UnlabeledImageSampleParser(...)

for sample in samples:
    sample_parser.with_sample(sample)
    img = sample_parser.get_image()
    if sample_parser.has_image_path:
        image_path = sample_parser.get_image_path()

    if sample_parser.has_image_metadata:
        image_metadata = sample_parser.get_image_metadata()

Attributes:

has_image_path	Whether this parser produces paths to images on disk for samples that it parses.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
current_sample	The current sample.

Methods:

get_image()	Returns the image from the current sample.
get_image_path()	Returns the image path for the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
clear_sample()	Clears the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.UnlabeledVideoSampleParser

Bases: SampleParser

Interface for SampleParser instances that parse unlabeled video samples.

The general recipe for using UnlabeledVideoSampleParser instances is as follows:

sample_parser = UnlabeledVideoSampleParser(...)

for sample in samples:
    sample_parser.with_sample(sample)
    video_path = sample_parser.get_video_path()
    video_metadata = sample_parser.get_video_metadata()

Attributes:

has_video_metadata	Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.
current_sample	The current sample.

Methods:

get_video_path()	Returns the video path for the current sample.
get_video_metadata()	Returns the video metadata for the current sample.
clear_sample()	Clears the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_video_metadata

Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.

get_video_path()

Returns the video path for the current sample.

Returns:

the path to the video on disk

get_video_metadata()

Returns the video metadata for the current sample.

Returns:

a fiftyone.core.metadata.VideoMetadata instance

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.UnlabeledMediaSampleParser

Bases: SampleParser

Interface for SampleParser instances that parse unlabeled media samples.

The general recipe for using UnlabeledMediaSampleParser instances is as follows:

sample_parser = UnlabeledMediaSampleParser(...)

for sample in samples:
    sample_parser.with_sample(sample)
    filepath = sample_parser.get_media_path()
    metadata = sample_parser.get_metadata()

Attributes:

has_metadata	Whether this parser produces fiftyone.core.metadata.Metadata instances for samples that it parses.
current_sample	The current sample.

Methods:

get_media_path()	Returns the media path for the current sample.
get_metadata()	Returns the metadata for the current sample.
clear_sample()	Clears the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_metadata

Whether this parser produces fiftyone.core.metadata.Metadata instances for samples that it parses.

get_media_path()

Returns the media path for the current sample.

Returns:

the path to the media on disk

get_metadata()

Returns the metadata for the current sample.

Returns:

a fiftyone.core.metadata.Metadata instance

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.ImageSampleParser

Bases: UnlabeledImageSampleParser

Sample parser that parses unlabeled image samples.

This implementation assumes that the provided sample is either an image that can be converted to numpy format via np.asarray() or the path to an image on disk.

Attributes:

has_image_path	Whether this parser produces paths to images on disk for samples that it parses.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
current_sample	The current sample.

Methods:

get_image()	Returns the image from the current sample.
get_image_path()	Returns the image path for the current sample.
clear_sample()	Clears the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.VideoSampleParser

Bases: UnlabeledVideoSampleParser

Sample parser that parses unlabeled video samples.

This implementation assumes that the provided sample is a path to a video on disk.

Attributes:

has_video_metadata	Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.
current_sample	The current sample.

Methods:

get_video_path()	Returns the video path for the current sample.
clear_sample()	Clears the current sample.
get_video_metadata()	Returns the video metadata for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_video_metadata

Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.

get_video_path()

Returns the video path for the current sample.

Returns:

the path to the video on disk

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

get_video_metadata()

Returns the video metadata for the current sample.

Returns:

a fiftyone.core.metadata.VideoMetadata instance

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.MediaSampleParser

Bases: UnlabeledMediaSampleParser

Sample parser that parses unlabeled media samples.

This implementation assumes that the provided sample is a path to a media file on disk.

Attributes:

has_metadata	Whether this parser produces fiftyone.core.metadata.Metadata instances for samples that it parses.
current_sample	The current sample.

Methods:

get_media_path()	Returns the media path for the current sample.
clear_sample()	Clears the current sample.
get_metadata()	Returns the metadata for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_metadata

Whether this parser produces fiftyone.core.metadata.Metadata instances for samples that it parses.

get_media_path()

Returns the media path for the current sample.

Returns:

the path to the media on disk

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

get_metadata()

Returns the metadata for the current sample.

Returns:

a fiftyone.core.metadata.Metadata instance

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.LabeledImageSampleParser

Bases: SampleParser

Interface for SampleParser instances that parse labeled image samples.

Instances of this class must return images in numpy format and labels as fiftyone.core.labels.Label instances.

The general recipe for using LabeledImageSampleParser instances is as follows:

sample_parser = LabeledImageSampleParser(...)

for sample in samples:
    sample_parser.with_sample(sample)
    img = sample_parser.get_image()
    label = sample_parser.get_label()

    if sample_parser.has_image_path:
        image_path = sample_parser.get_image_path()

    if sample_parser.has_image_metadata:
        image_metadata = sample_parser.get_image_metadata()

Attributes:

has_image_path	Whether this parser produces paths to images on disk for samples that it parses.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
label_cls	The fiftyone.core.labels.Label class(es) returned by this parser.
current_sample	The current sample.

Methods:

get_image()	Returns the image from the current sample.
get_image_path()	Returns the image path for the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
get_label()	Returns the label for the current sample.
clear_sample()	Clears the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the labels that it may return

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_label()

Returns the label for the current sample.

Returns:

a fiftyone.core.labels.Label instance, or a dictionary mapping field names to fiftyone.core.labels.Label instances, or None if the sample is unlabeled

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.LabeledVideoSampleParser

Bases: SampleParser

Interface for SampleParser instances that parse labeled video samples.

The general recipe for using LabeledVideoSampleParser instances is as follows:

sample_parser = LabeledVideoSampleParser(...)

for sample in samples:
    sample_parser.with_sample(sample)
    video_path = sample_parser.get_video_path()
    label = sample_parser.get_label()
    frames = sample_parser.get_frame_labels()

    if sample_parser.has_video_metadata:
        video_metadata = sample_parser.get_video_metadata()

Attributes:

has_video_metadata	Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.
label_cls	The fiftyone.core.labels.Label class(es) returned by this parser within the sample-level labels that it produces.
frame_labels_cls	The fiftyone.core.labels.Label class(es) returned by this parser within the frame labels that it produces.
current_sample	The current sample.

Methods:

get_video_path()	Returns the video path for the current sample.
get_video_metadata()	Returns the video metadata for the current sample.
get_label()	Returns the sample-level labels for the current sample.
get_frame_labels()	Returns the frame labels for the current sample.
clear_sample()	Clears the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_video_metadata

Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser within the sample-level labels that it produces.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return sample-level labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single sample-level label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return sample-level label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the sample-level labels that it may return

property frame_labels_cls

The fiftyone.core.labels.Label class(es) returned by this parser within the frame labels that it produces.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return frame labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single frame label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return frame label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in each frame

• None. In this case, the parser makes no guarantees about the frame labels that it may return

get_video_path()

Returns the video path for the current sample.

Returns:

the path to the video on disk

get_video_metadata()

Returns the video metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_label()

Returns the sample-level labels for the current sample.

Returns:

a fiftyone.core.labels.Label instance, or a dictionary mapping field names to fiftyone.core.labels.Label instances, or None if the sample has no sample-level labels

get_frame_labels()

Returns the frame labels for the current sample.

Returns:

a dictionary mapping frame numbers to dictionaries that map label fields to fiftyone.core.labels.Label instances for each video frame, or None if the sample has no frame labels

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.LabeledImageTupleSampleParser

Bases: LabeledImageSampleParser

Generic sample parser that parses samples that are (image_or_path, label) tuples, where:

• image_or_path is either an image that can be converted to numpy format via np.asarray() or the path to an image on disk

• label is a fiftyone.core.labels.Label instance

This implementation provides a _current_image() property that caches the image for the current sample, for efficiency in case multiple getters require access to the image (e.g., to normalize coordinates, compute metadata, etc).

See the following subclasses of this parser for implementations that parse labels for common tasks:

• Image classification: ImageClassificationSampleParser

• Object detection: ImageDetectionSampleParser

• Multitask image prediction: ImageLabelsSampleParser

Attributes:

has_image_path	Whether this parser produces paths to images on disk for samples that it parses.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
label_cls	The fiftyone.core.labels.Label class(es) returned by this parser.
current_sample	The current sample.

Methods:

get_image()	Returns the image from the current sample.
get_image_path()	Returns the image path for the current sample.
get_label()	Returns the label for the current sample.
clear_sample()	Clears the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the labels that it may return

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

get_label()

Returns the label for the current sample.

Returns:

a fiftyone.core.labels.Label instance, or a dictionary mapping field names to fiftyone.core.labels.Label instances, or None if the sample is unlabeled

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.ImageClassificationSampleParser(classes=None)

Bases: LabeledImageTupleSampleParser

Generic parser for image classification(s) samples whose labels are represented as fiftyone.core.labels.Classification instances.

This implementation supports samples that are (image_or_path, target) tuples, where:

• image_or_path is either an image that can be converted to numpy format via np.asarray() or the path to an image on disk

• target can be any of the following:

  • None, for unlabeled images

  • a label string or list of label strings

  • a class ID or list of class IDs, if classes is provided

  • a dict or list of dicts of the following form:

    {
        "label": <label-or-target>,
        "confidence": <confidence>,
        "attributes": <optional-attributes>,
    }
    

  • a fiftyone.core.labels.Classification or fiftyone.core.labels.Classifications instance

Parameters:

classes (None) – an optional list of class label strings. If provided, it is assumed that target contains class ID that should be mapped to label strings via classes[target]

Attributes:

label_cls	The fiftyone.core.labels.Label class(es) returned by this parser.
current_sample	The current sample.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
has_image_path	Whether this parser produces paths to images on disk for samples that it parses.

Methods:

get_label()	Returns the label for the current sample.
clear_sample()	Clears the current sample.
get_image()	Returns the image from the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
get_image_path()	Returns the image path for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the labels that it may return

get_label()

Returns the label for the current sample.

Parameters:

sample – the sample

Returns:

a fiftyone.core.labels.Classification instance

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.ImageDetectionSampleParser(label_field='label', bounding_box_field='bounding_box', confidence_field=None, attributes_field=None, classes=None, normalized=True)

Bases: LabeledImageTupleSampleParser

Generic parser for image detection samples whose labels are represented as fiftyone.core.labels.Detections instances.

This implementation supports samples that are (image_or_path, detections_or_path) tuples, where:

• image_or_path is either an image that can be converted to numpy format via np.asarray() or the path to an image on disk

• detections_or_path can be any of the following:

  • None, for unlabeled images

  • a list of detections in the following format:

    [
        {
            "<label_field>": <label-or-target>,
            "<bounding_box_field>": [
                <top-left-x>, <top-left-y>, <width>, <height>
            ],
            "<confidence_field>": <optional-confidence>,
            "<attributes_field>": {
                <optional-name>: <optional-value>,
                ...
            }
        },
        ...
    ]
    

    In the above, label-or-target is either a class ID (if classes is provided) or a label string, and the bounding box coordinates can either be relative coordinates in [0, 1] (if normalized == True) or absolute pixels coordinates (if normalized == False). The confidence and attributes fields are optional for each sample.

    The input field names can be configured as necessary when instantiating the parser.

  • the path on disk to a file in the above format

  • a fiftyone.core.labels.Detections instance

Parameters:

• label_field ("label") – the name of the object label field in the target dicts

• bounding_box_field ("bounding_box") – the name of the bounding box field in the target dicts

• confidence_field (None) – the name of the optional confidence field in the target dicts

• attributes_field (None) – the name of the optional attributes field in the target dicts

• classes (None) – an optional list of class label strings. If provided, it is assumed that the target values are class IDs that should be mapped to label strings via classes[target]

• normalized (True) – whether the bounding box coordinates are absolute pixel coordinates (False) or relative coordinates in [0, 1] (True)

Attributes:

label_cls	The fiftyone.core.labels.Label class(es) returned by this parser.
current_sample	The current sample.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
has_image_path	Whether this parser produces paths to images on disk for samples that it parses.

Methods:

get_label()	Returns the label for the current sample.
clear_sample()	Clears the current sample.
get_image()	Returns the image from the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
get_image_path()	Returns the image path for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the labels that it may return

get_label()

Returns the label for the current sample.

Returns:

a fiftyone.core.labels.Detections instance

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.ImageLabelsSampleParser(prefix=None, labels_dict=None, multilabel=False, skip_non_categorical=False)

Bases: LabeledImageTupleSampleParser

Generic parser for multitask image prediction samples whose labels are stored in eta.core.image.ImageLabels format.

This implementation provided by this class supports samples that are (image_or_path, image_labels_or_path) tuples, where:

• image_or_path is either an image that can be converted to numpy format via np.asarray() or the path to an image on disk

• image_labels_or_path is an eta.core.image.ImageLabels instance, an eta.core.frames.FrameLabels instance, a serialized dict representation of either, or the path to either on disk

Parameters:

• prefix (None) – a string prefix to prepend to each label name in the expanded label dictionary

• labels_dict (None) – a dictionary mapping names of attributes/objects in the image labels to field names into which to expand them

• multilabel (False) – whether to store attributes in a single fiftyone.core.labels.Classifications instance

• skip_non_categorical (False) – whether to skip non-categorical attributes (True) or cast them to strings (False)

Attributes:

label_cls	The fiftyone.core.labels.Label class(es) returned by this parser.
current_sample	The current sample.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
has_image_path	Whether this parser produces paths to images on disk for samples that it parses.

Methods:

get_label()	Returns the label for the current sample.
clear_sample()	Clears the current sample.
get_image()	Returns the image from the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
get_image_path()	Returns the image path for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the labels that it may return

get_label()

Returns the label for the current sample.

Returns:

a labels dictionary

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.FiftyOneImageClassificationSampleParser(classes=None)

Bases: ImageClassificationSampleParser

Parser for samples in FiftyOne image classification datasets.

See this page for format details.

Parameters:

classes (None) – an optional list of class label strings. If provided, it is assumed that target is a class ID that should be mapped to a label string via classes[target]

Methods:

clear_sample()	Clears the current sample.
get_image()	Returns the image from the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
get_image_path()	Returns the image path for the current sample.
get_label()	Returns the label for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Attributes:

current_sample	The current sample.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
has_image_path	Whether this parser produces paths to images on disk for samples that it parses.
label_cls	The fiftyone.core.labels.Label class(es) returned by this parser.

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

get_label()

Returns the label for the current sample.

Parameters:

sample – the sample

Returns:

a fiftyone.core.labels.Classification instance

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the labels that it may return

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.FiftyOneTemporalDetectionSampleParser(classes=None, compute_metadata=False)

Bases: LabeledVideoSampleParser

Parser for samples in FiftyOne temporal detection datasets.

See this page for format details.

Parameters:

• classes (None) – an optional list of class label strings. If provided, it is assumed that target is a class ID that should be mapped to a label string via classes[target]

• compute_metadata (False) – whether to compute fiftyone.core.metadata.VideoMetadata instances on-the-fly if get_video_metadata() is called and no metadata is available

Attributes:

has_video_metadata	Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.
label_cls	The fiftyone.core.labels.Label class(es) returned by this parser within the sample-level labels that it produces.
frame_labels_cls	The fiftyone.core.labels.Label class(es) returned by this parser within the frame labels that it produces.
current_sample	The current sample.

Methods:

with_sample(sample[, metadata])	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.
get_video_path()	Returns the video path for the current sample.
get_video_metadata()	Returns the video metadata for the current sample.
get_label()	Returns the sample-level labels for the current sample.
get_frame_labels()	Returns the frame labels for the current sample.
clear_sample()	Clears the current sample.

property has_video_metadata

Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser within the sample-level labels that it produces.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return sample-level labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single sample-level label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return sample-level label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the sample-level labels that it may return

property frame_labels_cls

The fiftyone.core.labels.Label class(es) returned by this parser within the frame labels that it produces.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return frame labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single frame label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return frame label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in each frame

• None. In this case, the parser makes no guarantees about the frame labels that it may return

with_sample(sample, metadata=None)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

get_video_path()

Returns the video path for the current sample.

Returns:

the path to the video on disk

get_video_metadata()

Returns the video metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_label()

Returns the sample-level labels for the current sample.

Returns:

a fiftyone.core.labels.Label instance, or a dictionary mapping field names to fiftyone.core.labels.Label instances, or None if the sample has no sample-level labels

get_frame_labels()

Returns the frame labels for the current sample.

Returns:

a dictionary mapping frame numbers to dictionaries that map label fields to fiftyone.core.labels.Label instances for each video frame, or None if the sample has no frame labels

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

class fiftyone.utils.data.parsers.FiftyOneImageDetectionSampleParser(classes=None)

Bases: ImageDetectionSampleParser

Parser for samples in FiftyOne image detection datasets.

See this page for format details.

Parameters:

classes (None) – an optional list of class label strings. If provided, it is assumed that the target values are class IDs that should be mapped to label strings via classes[target]

Methods:

clear_sample()	Clears the current sample.
get_image()	Returns the image from the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
get_image_path()	Returns the image path for the current sample.
get_label()	Returns the label for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Attributes:

current_sample	The current sample.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
has_image_path	Whether this parser produces paths to images on disk for samples that it parses.
label_cls	The fiftyone.core.labels.Label class(es) returned by this parser.

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

get_label()

Returns the label for the current sample.

Returns:

a fiftyone.core.labels.Detections instance

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the labels that it may return

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.FiftyOneImageLabelsSampleParser(prefix=None, labels_dict=None, multilabel=False, skip_non_categorical=False)

Bases: ImageLabelsSampleParser

Parser for samples in FiftyOne image labels datasets.

See this page for format details.

Parameters:

• prefix (None) – a string prefix to prepend to each label name in the expanded label dictionary

• labels_dict (None) – a dictionary mapping names of attributes/objects in the image labels to field names into which to expand them

• multilabel (False) – whether to store attributes in a single fiftyone.core.labels.Classifications instance

• skip_non_categorical (False) – whether to skip non-categorical attributes (True) or cast them to strings (False)

Methods:

clear_sample()	Clears the current sample.
get_image()	Returns the image from the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
get_image_path()	Returns the image path for the current sample.
get_label()	Returns the label for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Attributes:

current_sample	The current sample.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
has_image_path	Whether this parser produces paths to images on disk for samples that it parses.
label_cls	The fiftyone.core.labels.Label class(es) returned by this parser.

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

get_label()

Returns the label for the current sample.

Returns:

a labels dictionary

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the labels that it may return

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.VideoLabelsSampleParser(prefix=None, labels_dict=None, frame_labels_dict=None, multilabel=False, skip_non_categorical=False)

Bases: LabeledVideoSampleParser

Generic parser for labeled video samples whose labels are represented in eta.core.video.VideoLabels format.

This implementation provided by this class supports samples that are (video_path, video_labels_or_path) tuples, where:

• video_path is the path to a video on disk

• video_labels_or_path is an eta.core.video.VideoLabels instance, a serialized dict representation of one, or the path to one on disk

Parameters:

• prefix (None) – a string prefix to prepend to each label name in the expanded sample/frame label dictionaries

• labels_dict (None) – a dictionary mapping names of attributes/objects in the sample labels to field names into which to expand them. By default, all sample labels are loaded

• frame_labels_dict (None) – a dictionary mapping names of attributes/objects in the frame labels to field names into which to expand them. By default, all frame labels are loaded

• multilabel (False) – whether to store attributes in a single fiftyone.core.labels.Classifications instance

• skip_non_categorical (False) – whether to skip non-categorical attributes (True) or cast them to strings (False)

Attributes:

has_video_metadata	Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.
label_cls	The fiftyone.core.labels.Label class(es) returned by this parser within the sample-level labels that it produces.
frame_labels_cls	The fiftyone.core.labels.Label class(es) returned by this parser within the frame labels that it produces.
current_sample	The current sample.

Methods:

get_video_path()	Returns the video path for the current sample.
get_label()	Returns the sample-level labels for the current sample.
get_frame_labels()	Returns the frame labels for the current sample.
clear_sample()	Clears the current sample.
get_video_metadata()	Returns the video metadata for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_video_metadata

Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser within the sample-level labels that it produces.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return sample-level labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single sample-level label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return sample-level label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the sample-level labels that it may return

property frame_labels_cls

The fiftyone.core.labels.Label class(es) returned by this parser within the frame labels that it produces.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return frame labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single frame label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return frame label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in each frame

• None. In this case, the parser makes no guarantees about the frame labels that it may return

get_video_path()

Returns the video path for the current sample.

Returns:

the path to the video on disk

get_label()

Returns the sample-level labels for the current sample.

Returns:

a fiftyone.core.labels.Label instance, or a dictionary mapping field names to fiftyone.core.labels.Label instances, or None if the sample has no sample-level labels

get_frame_labels()

Returns the frame labels for the current sample.

Returns:

a dictionary mapping frame numbers to dictionaries that map label fields to fiftyone.core.labels.Label instances for each video frame, or None if the sample has no frame labels

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

get_video_metadata()

Returns the video metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.FiftyOneVideoLabelsSampleParser(prefix=None, labels_dict=None, frame_labels_dict=None, multilabel=False, skip_non_categorical=False)

Bases: VideoLabelsSampleParser

Parser for samples in FiftyOne video labels datasets.

See this page for format details.

Parameters:

• expand (True) – whether to expand the labels for each frame into separate fiftyone.core.labels.Label instances

• prefix (None) – a string prefix to prepend to each label name in the expanded frame label dictionaries

• labels_dict (None) – a dictionary mapping names of attributes/objects in the frame labels to field names into which to expand them

• multilabel (False) – whether to store attributes in a single fiftyone.core.labels.Classifications instance

• skip_non_categorical (False) – whether to skip non-categorical attributes (True) or cast them to strings (False)

Methods:

clear_sample()	Clears the current sample.
get_frame_labels()	Returns the frame labels for the current sample.
get_label()	Returns the sample-level labels for the current sample.
get_video_metadata()	Returns the video metadata for the current sample.
get_video_path()	Returns the video path for the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Attributes:

current_sample	The current sample.
frame_labels_cls	The fiftyone.core.labels.Label class(es) returned by this parser within the frame labels that it produces.
has_video_metadata	Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.
label_cls	The fiftyone.core.labels.Label class(es) returned by this parser within the sample-level labels that it produces.

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

property frame_labels_cls

The fiftyone.core.labels.Label class(es) returned by this parser within the frame labels that it produces.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return frame labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single frame label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return frame label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in each frame

• None. In this case, the parser makes no guarantees about the frame labels that it may return

get_frame_labels()

Returns the frame labels for the current sample.

Returns:

a dictionary mapping frame numbers to dictionaries that map label fields to fiftyone.core.labels.Label instances for each video frame, or None if the sample has no frame labels

get_label()

Returns the sample-level labels for the current sample.

Returns:

a fiftyone.core.labels.Label instance, or a dictionary mapping field names to fiftyone.core.labels.Label instances, or None if the sample has no sample-level labels

get_video_metadata()

Returns the video metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_video_path()

Returns the video path for the current sample.

Returns:

the path to the video on disk

property has_video_metadata

Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser within the sample-level labels that it produces.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return sample-level labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single sample-level label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return sample-level label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the sample-level labels that it may return

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.FiftyOneUnlabeledImageSampleParser(compute_metadata=False)

Bases: UnlabeledImageSampleParser

Parser for fiftyone.core.sample.Sample instances that contain images.

Parameters:

compute_metadata (False) – whether to compute fiftyone.core.metadata.ImageMetadata instances on-the-fly if get_image_metadata() is called and no metadata is available

Attributes:

has_image_path	Whether this parser produces paths to images on disk for samples that it parses.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
current_sample	The current sample.

Methods:

get_image()	Returns the image from the current sample.
get_image_path()	Returns the image path for the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
clear_sample()	Clears the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.FiftyOneLabeledImageSampleParser(label_field, label_fcn=None, compute_metadata=False)

Bases: LabeledImageSampleParser

Parser for fiftyone.core.sample.Sample instances that contain labeled images.

Parameters:

• label_field – the name of the label field to parse, or a dictionary mapping label field names to keys for the return label dictionaries

• label_fcn (None) – an optional function or dictionary mapping label field names to functions (must match label_field) to apply to each label before returning it

• compute_metadata (False) – whether to compute fiftyone.core.metadata.ImageMetadata instances on-the-fly if get_image_metadata() is called and no metadata is available

Attributes:

has_image_path	Whether this parser produces paths to images on disk for samples that it parses.
has_image_metadata	Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.
label_cls	The fiftyone.core.labels.Label class(es) returned by this parser.
current_sample	The current sample.

Methods:

get_image()	Returns the image from the current sample.
get_image_path()	Returns the image path for the current sample.
get_image_metadata()	Returns the image metadata for the current sample.
get_label()	Returns the label for the current sample.
clear_sample()	Clears the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_image_path

Whether this parser produces paths to images on disk for samples that it parses.

property has_image_metadata

Whether this parser produces fiftyone.core.metadata.ImageMetadata instances for samples that it parses.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the labels that it may return

get_image()

Returns the image from the current sample.

Returns:

a numpy image

get_image_path()

Returns the image path for the current sample.

Returns:

the path to the image on disk

get_image_metadata()

Returns the image metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_label()

Returns the label for the current sample.

Returns:

a fiftyone.core.labels.Label instance, or a dictionary mapping field names to fiftyone.core.labels.Label instances, or None if the sample is unlabeled

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.ExtractClipsMixin(compute_metadata=False, write_clips=True, clip_dir=None, video_format=None)

Bases: object

Mixin for sample parsers that extract clips from fiftyone.core.clips.ClipView instances.

Parameters:

• compute_metadata (False) – whether to compute fiftyone.core.metadata.VideoMetadata instances on-the-fly when no pre-computed metadata is available

• write_clips (True) – whether to write clips when their paths are requested

• clip_dir (None) – a directory to write clips. Only applicable when parsing fiftyone.core.clips.ClipView instances

• video_format (None) – the video format to use when writing video clips to disk. By default, fiftyone.config.default_video_ext is used

class fiftyone.utils.data.parsers.FiftyOneUnlabeledVideoSampleParser(compute_metadata=False, write_clips=True, clip_dir=None, video_format=None)

Bases: ExtractClipsMixin, UnlabeledVideoSampleParser

Parser for fiftyone.core.sample.Sample instances that contain videos.

This class also supports fiftyone.core.clips.ClipView instances.

Parameters:

• compute_metadata (False) – whether to compute fiftyone.core.metadata.VideoMetadata instances on-the-fly if get_video_metadata() is called and no metadata is available

• write_clips (True) – whether to write clips when get_video_path() is called

• clip_dir (None) – a directory to write clips. Only applicable when parsing fiftyone.core.clips.ClipView instances

• video_format (None) – the video format to use when writing video clips to disk. By default, fiftyone.config.default_video_ext is used

Attributes:

has_video_metadata	Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.
current_sample	The current sample.

Methods:

get_video_path()	Returns the video path for the current sample.
get_video_metadata()	Returns the video metadata for the current sample.
clear_sample()	Clears the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_video_metadata

Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.

get_video_path()

Returns the video path for the current sample.

Returns:

the path to the video on disk

get_video_metadata()

Returns the video metadata for the current sample.

Returns:

a fiftyone.core.metadata.VideoMetadata instance

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.FiftyOneLabeledVideoSampleParser(label_field=None, frame_labels_field=None, label_fcn=None, frame_labels_fcn=None, compute_metadata=False, write_clips=True, clip_dir=None, video_format=None)

Bases: ExtractClipsMixin, LabeledVideoSampleParser

Parser for fiftyone.core.sample.Sample instances that contain labeled videos.

This class also supports fiftyone.core.clips.ClipView instances.

Parameters:

• label_field (None) – the name of a label field to parse, or a dictionary mapping label field names to output keys to use in the returned sample-level labels dictionary

• frame_labels_field (None) – the name of a frame label field to parse, or a dictionary mapping field names to output keys describing the frame label fields to export

• label_fcn (None) – an optional function or dictionary mapping label field names to functions (must match label_field) to apply to each sample label before returning it

• frame_labels_fcn (None) – an optional function or dictionary mapping frame label field names to functions (must match frame_labels_field) to apply to each frame label before returning it

• compute_metadata (False) – whether to compute fiftyone.core.metadata.VideoMetadata instances on-the-fly if get_video_metadata() is called and no metadata is available

• write_clips (True) – whether to write clips when get_video_path() is called

• clip_dir (None) – a directory to write clips. Only applicable when parsing fiftyone.core.clips.ClipView instances

• video_format (None) – the video format to use when writing video clips to disk. By default, fiftyone.config.default_video_ext is used

Attributes:

has_video_metadata	Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.
label_cls	The fiftyone.core.labels.Label class(es) returned by this parser within the sample-level labels that it produces.
frame_labels_cls	The fiftyone.core.labels.Label class(es) returned by this parser within the frame labels that it produces.
current_sample	The current sample.

Methods:

get_video_path()	Returns the video path for the current sample.
get_video_metadata()	Returns the video metadata for the current sample.
get_label()	Returns the sample-level labels for the current sample.
get_frame_labels()	Returns the frame labels for the current sample.
clear_sample()	Clears the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_video_metadata

Whether this parser produces fiftyone.core.metadata.VideoMetadata instances for samples that it parses.

property label_cls

The fiftyone.core.labels.Label class(es) returned by this parser within the sample-level labels that it produces.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return sample-level labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single sample-level label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return sample-level label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in the imported labels

• None. In this case, the parser makes no guarantees about the sample-level labels that it may return

property frame_labels_cls

The fiftyone.core.labels.Label class(es) returned by this parser within the frame labels that it produces.

This can be any of the following:

• a fiftyone.core.labels.Label class. In this case, the parser is guaranteed to return frame labels of this type

• a list or tuple of fiftyone.core.labels.Label classes. In this case, the parser can produce a single frame label field of any of these types

• a dict mapping keys to fiftyone.core.labels.Label classes. In this case, the parser will return frame label dictionaries with keys and value-types specified by this dictionary. Not all keys need be present in each frame

• None. In this case, the parser makes no guarantees about the frame labels that it may return

get_video_path()

Returns the video path for the current sample.

Returns:

the path to the video on disk

get_video_metadata()

Returns the video metadata for the current sample.

Returns:

a fiftyone.core.metadata.ImageMetadata instance

get_label()

Returns the sample-level labels for the current sample.

Returns:

a fiftyone.core.labels.Label instance, or a dictionary mapping field names to fiftyone.core.labels.Label instances, or None if the sample has no sample-level labels

get_frame_labels()

Returns the frame labels for the current sample.

Returns:

a dictionary mapping frame numbers to dictionaries that map label fields to fiftyone.core.labels.Label instances for each video frame, or None if the sample has no frame labels

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample

class fiftyone.utils.data.parsers.FiftyOneUnlabeledMediaSampleParser(compute_metadata=False)

Bases: MediaSampleParser

Parser for fiftyone.core.sample.Sample instances that contain unlabeled media.

Parameters:

compute_metadata (False) – whether to compute fiftyone.core.metadata.Metadata instances on-the-fly if get_metadata() is called and no metadata is available

Attributes:

has_metadata	Whether this parser produces fiftyone.core.metadata.Metadata instances for samples that it parses.
current_sample	The current sample.

Methods:

get_media_path()	Returns the media path for the current sample.
get_metadata()	Returns the metadata for the current sample.
clear_sample()	Clears the current sample.
with_sample(sample)	Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

property has_metadata

Whether this parser produces fiftyone.core.metadata.Metadata instances for samples that it parses.

get_media_path()

Returns the media path for the current sample.

Returns:

the path to the media on disk

get_metadata()

Returns the metadata for the current sample.

Returns:

a fiftyone.core.metadata.Metadata instance

clear_sample()

Clears the current sample.

Also clears any cached sample information stored by the parser.

property current_sample

The current sample.

Raises:

ValueError – if there is no current sample

with_sample(sample)

Sets the current sample so that subsequent calls to methods of this parser will return information from the given sample.

Guaranteed to call clear_sample() before setting the current sample.

Parameters:

sample – a sample