Skip to content

Datasets

Warning

Dataset API is still fluid and may change. If you use Dataset API in your project until further notice, freeze the supervision version in your requirements.txt or setup.py.

DetectionDataset

Bases: BaseDataset

Dataclass containing information about object detection dataset.

Attributes:

Name Type Description
classes List[str]

List containing dataset class names.

images Dict[str, ndarray]

Dictionary mapping image name to image.

annotations Dict[str, Detections]

Dictionary mapping image name to annotations.

Source code in supervision/dataset/core.py
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
@dataclass
class DetectionDataset(BaseDataset):
    """
    Dataclass containing information about object detection dataset.

    Attributes:
        classes (List[str]): List containing dataset class names.
        images (Dict[str, np.ndarray]): Dictionary mapping image name to image.
        annotations (Dict[str, Detections]): Dictionary mapping
            image name to annotations.
    """

    classes: List[str]
    images: Dict[str, np.ndarray]
    annotations: Dict[str, Detections]

    def __len__(self) -> int:
        """
        Return the number of images in the dataset.

        Returns:
            int: The number of images.
        """
        return len(self.images)

    def __iter__(self) -> Iterator[Tuple[str, np.ndarray, Detections]]:
        """
        Iterate over the images and annotations in the dataset.

        Yields:
            Iterator[Tuple[str, np.ndarray, Detections]]:
                An iterator that yields tuples containing the image name,
                the image data, and its corresponding annotation.
        """
        for image_name, image in self.images.items():
            yield image_name, image, self.annotations.get(image_name, None)

    def __eq__(self, other):
        if not isinstance(other, DetectionDataset):
            return False

        if set(self.classes) != set(other.classes):
            return False

        for key in self.images:
            if not np.array_equal(self.images[key], other.images[key]):
                return False
            if not self.annotations[key] == other.annotations[key]:
                return False

        return True

    def split(
        self, split_ratio=0.8, random_state=None, shuffle: bool = True
    ) -> Tuple[DetectionDataset, DetectionDataset]:
        """
        Splits the dataset into two parts (training and testing)
            using the provided split_ratio.

        Args:
            split_ratio (float, optional): The ratio of the training
                set to the entire dataset.
            random_state (int, optional): The seed for the random number generator.
                This is used for reproducibility.
            shuffle (bool, optional): Whether to shuffle the data before splitting.

        Returns:
            Tuple[DetectionDataset, DetectionDataset]: A tuple containing
                the training and testing datasets.

        Examples:
            ```python
            import supervision as sv

            ds = sv.DetectionDataset(...)
            train_ds, test_ds = ds.split(split_ratio=0.7, random_state=42, shuffle=True)
            len(train_ds), len(test_ds)
            # (700, 300)
            ```
        """

        image_names = list(self.images.keys())
        train_names, test_names = train_test_split(
            data=image_names,
            train_ratio=split_ratio,
            random_state=random_state,
            shuffle=shuffle,
        )

        train_dataset = DetectionDataset(
            classes=self.classes,
            images={name: self.images[name] for name in train_names},
            annotations={name: self.annotations[name] for name in train_names},
        )
        test_dataset = DetectionDataset(
            classes=self.classes,
            images={name: self.images[name] for name in test_names},
            annotations={name: self.annotations[name] for name in test_names},
        )
        return train_dataset, test_dataset

    def as_pascal_voc(
        self,
        images_directory_path: Optional[str] = None,
        annotations_directory_path: Optional[str] = None,
        min_image_area_percentage: float = 0.0,
        max_image_area_percentage: float = 1.0,
        approximation_percentage: float = 0.0,
    ) -> None:
        """
        Exports the dataset to PASCAL VOC format. This method saves the images
        and their corresponding annotations in PASCAL VOC format.

        Args:
            images_directory_path (Optional[str]): The path to the directory
                where the images should be saved.
                If not provided, images will not be saved.
            annotations_directory_path (Optional[str]): The path to
                the directory where the annotations in PASCAL VOC format should be
                saved. If not provided, annotations will not be saved.
            min_image_area_percentage (float): The minimum percentage of
                detection area relative to
                the image area for a detection to be included.
                Argument is used only for segmentation datasets.
            max_image_area_percentage (float): The maximum percentage
                of detection area relative to
                the image area for a detection to be included.
                Argument is used only for segmentation datasets.
            approximation_percentage (float): The percentage of
                polygon points to be removed from the input polygon,
                in the range [0, 1). Argument is used only for segmentation datasets.
        """
        if images_directory_path:
            save_dataset_images(
                images_directory_path=images_directory_path, images=self.images
            )
        if annotations_directory_path:
            Path(annotations_directory_path).mkdir(parents=True, exist_ok=True)

        for image_path, image in self.images.items():
            detections = self.annotations[image_path]

            if annotations_directory_path:
                annotation_name = Path(image_path).stem
                annotations_path = os.path.join(
                    annotations_directory_path, f"{annotation_name}.xml"
                )
                image_name = Path(image_path).name
                pascal_voc_xml = detections_to_pascal_voc(
                    detections=detections,
                    classes=self.classes,
                    filename=image_name,
                    image_shape=image.shape,
                    min_image_area_percentage=min_image_area_percentage,
                    max_image_area_percentage=max_image_area_percentage,
                    approximation_percentage=approximation_percentage,
                )

                with open(annotations_path, "w") as f:
                    f.write(pascal_voc_xml)

    @classmethod
    def from_pascal_voc(
        cls,
        images_directory_path: str,
        annotations_directory_path: str,
        force_masks: bool = False,
    ) -> DetectionDataset:
        """
        Creates a Dataset instance from PASCAL VOC formatted data.

        Args:
            images_directory_path (str): Path to the directory containing the images.
            annotations_directory_path (str): Path to the directory
                containing the PASCAL VOC XML annotations.
            force_masks (bool, optional): If True, forces masks to
                be loaded for all annotations, regardless of whether they are present.

        Returns:
            DetectionDataset: A DetectionDataset instance containing
                the loaded images and annotations.

        Examples:
            ```python
            import roboflow
            from roboflow import Roboflow
            import supervision as sv

            roboflow.login()

            rf = Roboflow()

            project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
            dataset = project.version(PROJECT_VERSION).download("voc")

            ds = sv.DetectionDataset.from_pascal_voc(
                images_directory_path=f"{dataset.location}/train/images",
                annotations_directory_path=f"{dataset.location}/train/labels"
            )

            ds.classes
            # ['dog', 'person']
            ```
        """

        classes, images, annotations = load_pascal_voc_annotations(
            images_directory_path=images_directory_path,
            annotations_directory_path=annotations_directory_path,
            force_masks=force_masks,
        )

        return DetectionDataset(classes=classes, images=images, annotations=annotations)

    @classmethod
    def from_yolo(
        cls,
        images_directory_path: str,
        annotations_directory_path: str,
        data_yaml_path: str,
        force_masks: bool = False,
    ) -> DetectionDataset:
        """
        Creates a Dataset instance from YOLO formatted data.

        Args:
            images_directory_path (str): The path to the
                directory containing the images.
            annotations_directory_path (str): The path to the directory
                containing the YOLO annotation files.
            data_yaml_path (str): The path to the data
                YAML file containing class information.
            force_masks (bool, optional): If True, forces
                masks to be loaded for all annotations,
                regardless of whether they are present.

        Returns:
            DetectionDataset: A DetectionDataset instance
                containing the loaded images and annotations.

        Examples:
            ```python
            import roboflow
            from roboflow import Roboflow
            import supervision as sv

            roboflow.login()
            rf = Roboflow()

            project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
            dataset = project.version(PROJECT_VERSION).download("yolov5")

            ds = sv.DetectionDataset.from_yolo(
                images_directory_path=f"{dataset.location}/train/images",
                annotations_directory_path=f"{dataset.location}/train/labels",
                data_yaml_path=f"{dataset.location}/data.yaml"
            )

            ds.classes
            # ['dog', 'person']
            ```
        """
        classes, images, annotations = load_yolo_annotations(
            images_directory_path=images_directory_path,
            annotations_directory_path=annotations_directory_path,
            data_yaml_path=data_yaml_path,
            force_masks=force_masks,
        )
        return DetectionDataset(classes=classes, images=images, annotations=annotations)

    def as_yolo(
        self,
        images_directory_path: Optional[str] = None,
        annotations_directory_path: Optional[str] = None,
        data_yaml_path: Optional[str] = None,
        min_image_area_percentage: float = 0.0,
        max_image_area_percentage: float = 1.0,
        approximation_percentage: float = 0.0,
    ) -> None:
        """
        Exports the dataset to YOLO format. This method saves the
        images and their corresponding annotations in YOLO format.

        Args:
            images_directory_path (Optional[str]): The path to the
                directory where the images should be saved.
                If not provided, images will not be saved.
            annotations_directory_path (Optional[str]): The path to the
                directory where the annotations in
                YOLO format should be saved. If not provided,
                annotations will not be saved.
            data_yaml_path (Optional[str]): The path where the data.yaml
                file should be saved.
                If not provided, the file will not be saved.
            min_image_area_percentage (float): The minimum percentage of
                detection area relative to
                the image area for a detection to be included.
                Argument is used only for segmentation datasets.
            max_image_area_percentage (float): The maximum percentage
                of detection area relative to
                the image area for a detection to be included.
                Argument is used only for segmentation datasets.
            approximation_percentage (float): The percentage of polygon points to
                be removed from the input polygon, in the range [0, 1).
                This is useful for simplifying the annotations.
                Argument is used only for segmentation datasets.
        """
        if images_directory_path is not None:
            save_dataset_images(
                images_directory_path=images_directory_path, images=self.images
            )
        if annotations_directory_path is not None:
            save_yolo_annotations(
                annotations_directory_path=annotations_directory_path,
                images=self.images,
                annotations=self.annotations,
                min_image_area_percentage=min_image_area_percentage,
                max_image_area_percentage=max_image_area_percentage,
                approximation_percentage=approximation_percentage,
            )
        if data_yaml_path is not None:
            save_data_yaml(data_yaml_path=data_yaml_path, classes=self.classes)

    @classmethod
    def from_coco(
        cls,
        images_directory_path: str,
        annotations_path: str,
        force_masks: bool = False,
    ) -> DetectionDataset:
        """
        Creates a Dataset instance from COCO formatted data.

        Args:
            images_directory_path (str): The path to the
                directory containing the images.
            annotations_path (str): The path to the json annotation files.
            force_masks (bool, optional): If True,
                forces masks to be loaded for all annotations,
                regardless of whether they are present.

        Returns:
            DetectionDataset: A DetectionDataset instance containing
                the loaded images and annotations.

        Examples:
            ```python
            import roboflow
            from roboflow import Roboflow
            import supervision as sv

            roboflow.login()
            rf = Roboflow()

            project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
            dataset = project.version(PROJECT_VERSION).download("coco")

            ds = sv.DetectionDataset.from_coco(
                images_directory_path=f"{dataset.location}/train",
                annotations_path=f"{dataset.location}/train/_annotations.coco.json",
            )

            ds.classes
            # ['dog', 'person']
            ```
        """
        classes, images, annotations = load_coco_annotations(
            images_directory_path=images_directory_path,
            annotations_path=annotations_path,
            force_masks=force_masks,
        )
        return DetectionDataset(classes=classes, images=images, annotations=annotations)

    def as_coco(
        self,
        images_directory_path: Optional[str] = None,
        annotations_path: Optional[str] = None,
        min_image_area_percentage: float = 0.0,
        max_image_area_percentage: float = 1.0,
        approximation_percentage: float = 0.0,
    ) -> None:
        """
        Exports the dataset to COCO format. This method saves the
        images and their corresponding annotations in COCO format.

        !!! tip

            The format of the mask is determined automatically based on its structure:

            - If a mask contains multiple disconnected components or holes, it will be
            saved using the Run-Length Encoding (RLE) format for efficient storage and
            processing.
            - If a mask consists of a single, contiguous region without any holes, it
            will be encoded as a polygon, preserving the outline of the object.

            This automatic selection ensures that the masks are stored in the most
            appropriate and space-efficient format, complying with COCO dataset
            standards.

        Args:
            images_directory_path (Optional[str]): The path to the directory
                where the images should be saved.
                If not provided, images will not be saved.
            annotations_path (Optional[str]): The path to COCO annotation file.
            min_image_area_percentage (float): The minimum percentage of
                detection area relative to
                the image area for a detection to be included.
                Argument is used only for segmentation datasets.
            max_image_area_percentage (float): The maximum percentage of
                detection area relative to
                the image area for a detection to be included.
                Argument is used only for segmentation datasets.
            approximation_percentage (float): The percentage of polygon points
                to be removed from the input polygon,
                in the range [0, 1). This is useful for simplifying the annotations.
                Argument is used only for segmentation datasets.
        """
        if images_directory_path is not None:
            save_dataset_images(
                images_directory_path=images_directory_path, images=self.images
            )
        if annotations_path is not None:
            save_coco_annotations(
                annotation_path=annotations_path,
                images=self.images,
                annotations=self.annotations,
                classes=self.classes,
                min_image_area_percentage=min_image_area_percentage,
                max_image_area_percentage=max_image_area_percentage,
                approximation_percentage=approximation_percentage,
            )

    @classmethod
    def merge(cls, dataset_list: List[DetectionDataset]) -> DetectionDataset:
        """
        Merge a list of `DetectionDataset` objects into a single
            `DetectionDataset` object.

        This method takes a list of `DetectionDataset` objects and combines
        their respective fields (`classes`, `images`,
        `annotations`) into a single `DetectionDataset` object.

        Args:
            dataset_list (List[DetectionDataset]): A list of `DetectionDataset`
                objects to merge.

        Returns:
            (DetectionDataset): A single `DetectionDataset` object containing
            the merged data from the input list.

        Examples:
            ```python
            import supervision as sv

            ds_1 = sv.DetectionDataset(...)
            len(ds_1)
            # 100
            ds_1.classes
            # ['dog', 'person']

            ds_2 = sv.DetectionDataset(...)
            len(ds_2)
            # 200
            ds_2.classes
            # ['cat']

            ds_merged = sv.DetectionDataset.merge([ds_1, ds_2])
            len(ds_merged)
            # 300
            ds_merged.classes
            # ['cat', 'dog', 'person']
            ```
        """
        merged_images, merged_annotations = {}, {}
        class_lists = [dataset.classes for dataset in dataset_list]
        merged_classes = merge_class_lists(class_lists=class_lists)

        for dataset in dataset_list:
            class_index_mapping = build_class_index_mapping(
                source_classes=dataset.classes, target_classes=merged_classes
            )
            for image_name, image, detections in dataset:
                if image_name in merged_annotations:
                    raise ValueError(
                        f"Image name {image_name} is not unique across datasets."
                    )

                merged_images[image_name] = image
                merged_annotations[image_name] = map_detections_class_id(
                    source_to_target_mapping=class_index_mapping,
                    detections=detections,
                )

        return cls(
            classes=merged_classes, images=merged_images, annotations=merged_annotations
        )

Functions

__iter__()

Iterate over the images and annotations in the dataset.

Yields:

Type Description
str

Iterator[Tuple[str, np.ndarray, Detections]]: An iterator that yields tuples containing the image name, the image data, and its corresponding annotation.

Source code in supervision/dataset/core.py
def __iter__(self) -> Iterator[Tuple[str, np.ndarray, Detections]]:
    """
    Iterate over the images and annotations in the dataset.

    Yields:
        Iterator[Tuple[str, np.ndarray, Detections]]:
            An iterator that yields tuples containing the image name,
            the image data, and its corresponding annotation.
    """
    for image_name, image in self.images.items():
        yield image_name, image, self.annotations.get(image_name, None)

__len__()

Return the number of images in the dataset.

Returns:

Name Type Description
int int

The number of images.

Source code in supervision/dataset/core.py
def __len__(self) -> int:
    """
    Return the number of images in the dataset.

    Returns:
        int: The number of images.
    """
    return len(self.images)

as_coco(images_directory_path=None, annotations_path=None, min_image_area_percentage=0.0, max_image_area_percentage=1.0, approximation_percentage=0.0)

Exports the dataset to COCO format. This method saves the images and their corresponding annotations in COCO format.

Tip

The format of the mask is determined automatically based on its structure:

  • If a mask contains multiple disconnected components or holes, it will be saved using the Run-Length Encoding (RLE) format for efficient storage and processing.
  • If a mask consists of a single, contiguous region without any holes, it will be encoded as a polygon, preserving the outline of the object.

This automatic selection ensures that the masks are stored in the most appropriate and space-efficient format, complying with COCO dataset standards.

Parameters:

Name Type Description Default
images_directory_path Optional[str]

The path to the directory where the images should be saved. If not provided, images will not be saved.

None
annotations_path Optional[str]

The path to COCO annotation file.

None
min_image_area_percentage float

The minimum percentage of detection area relative to the image area for a detection to be included. Argument is used only for segmentation datasets.

0.0
max_image_area_percentage float

The maximum percentage of detection area relative to the image area for a detection to be included. Argument is used only for segmentation datasets.

1.0
approximation_percentage float

The percentage of polygon points to be removed from the input polygon, in the range [0, 1). This is useful for simplifying the annotations. Argument is used only for segmentation datasets.

0.0
Source code in supervision/dataset/core.py
def as_coco(
    self,
    images_directory_path: Optional[str] = None,
    annotations_path: Optional[str] = None,
    min_image_area_percentage: float = 0.0,
    max_image_area_percentage: float = 1.0,
    approximation_percentage: float = 0.0,
) -> None:
    """
    Exports the dataset to COCO format. This method saves the
    images and their corresponding annotations in COCO format.

    !!! tip

        The format of the mask is determined automatically based on its structure:

        - If a mask contains multiple disconnected components or holes, it will be
        saved using the Run-Length Encoding (RLE) format for efficient storage and
        processing.
        - If a mask consists of a single, contiguous region without any holes, it
        will be encoded as a polygon, preserving the outline of the object.

        This automatic selection ensures that the masks are stored in the most
        appropriate and space-efficient format, complying with COCO dataset
        standards.

    Args:
        images_directory_path (Optional[str]): The path to the directory
            where the images should be saved.
            If not provided, images will not be saved.
        annotations_path (Optional[str]): The path to COCO annotation file.
        min_image_area_percentage (float): The minimum percentage of
            detection area relative to
            the image area for a detection to be included.
            Argument is used only for segmentation datasets.
        max_image_area_percentage (float): The maximum percentage of
            detection area relative to
            the image area for a detection to be included.
            Argument is used only for segmentation datasets.
        approximation_percentage (float): The percentage of polygon points
            to be removed from the input polygon,
            in the range [0, 1). This is useful for simplifying the annotations.
            Argument is used only for segmentation datasets.
    """
    if images_directory_path is not None:
        save_dataset_images(
            images_directory_path=images_directory_path, images=self.images
        )
    if annotations_path is not None:
        save_coco_annotations(
            annotation_path=annotations_path,
            images=self.images,
            annotations=self.annotations,
            classes=self.classes,
            min_image_area_percentage=min_image_area_percentage,
            max_image_area_percentage=max_image_area_percentage,
            approximation_percentage=approximation_percentage,
        )

as_pascal_voc(images_directory_path=None, annotations_directory_path=None, min_image_area_percentage=0.0, max_image_area_percentage=1.0, approximation_percentage=0.0)

Exports the dataset to PASCAL VOC format. This method saves the images and their corresponding annotations in PASCAL VOC format.

Parameters:

Name Type Description Default
images_directory_path Optional[str]

The path to the directory where the images should be saved. If not provided, images will not be saved.

None
annotations_directory_path Optional[str]

The path to the directory where the annotations in PASCAL VOC format should be saved. If not provided, annotations will not be saved.

None
min_image_area_percentage float

The minimum percentage of detection area relative to the image area for a detection to be included. Argument is used only for segmentation datasets.

0.0
max_image_area_percentage float

The maximum percentage of detection area relative to the image area for a detection to be included. Argument is used only for segmentation datasets.

1.0
approximation_percentage float

The percentage of polygon points to be removed from the input polygon, in the range [0, 1). Argument is used only for segmentation datasets.

0.0
Source code in supervision/dataset/core.py
def as_pascal_voc(
    self,
    images_directory_path: Optional[str] = None,
    annotations_directory_path: Optional[str] = None,
    min_image_area_percentage: float = 0.0,
    max_image_area_percentage: float = 1.0,
    approximation_percentage: float = 0.0,
) -> None:
    """
    Exports the dataset to PASCAL VOC format. This method saves the images
    and their corresponding annotations in PASCAL VOC format.

    Args:
        images_directory_path (Optional[str]): The path to the directory
            where the images should be saved.
            If not provided, images will not be saved.
        annotations_directory_path (Optional[str]): The path to
            the directory where the annotations in PASCAL VOC format should be
            saved. If not provided, annotations will not be saved.
        min_image_area_percentage (float): The minimum percentage of
            detection area relative to
            the image area for a detection to be included.
            Argument is used only for segmentation datasets.
        max_image_area_percentage (float): The maximum percentage
            of detection area relative to
            the image area for a detection to be included.
            Argument is used only for segmentation datasets.
        approximation_percentage (float): The percentage of
            polygon points to be removed from the input polygon,
            in the range [0, 1). Argument is used only for segmentation datasets.
    """
    if images_directory_path:
        save_dataset_images(
            images_directory_path=images_directory_path, images=self.images
        )
    if annotations_directory_path:
        Path(annotations_directory_path).mkdir(parents=True, exist_ok=True)

    for image_path, image in self.images.items():
        detections = self.annotations[image_path]

        if annotations_directory_path:
            annotation_name = Path(image_path).stem
            annotations_path = os.path.join(
                annotations_directory_path, f"{annotation_name}.xml"
            )
            image_name = Path(image_path).name
            pascal_voc_xml = detections_to_pascal_voc(
                detections=detections,
                classes=self.classes,
                filename=image_name,
                image_shape=image.shape,
                min_image_area_percentage=min_image_area_percentage,
                max_image_area_percentage=max_image_area_percentage,
                approximation_percentage=approximation_percentage,
            )

            with open(annotations_path, "w") as f:
                f.write(pascal_voc_xml)

as_yolo(images_directory_path=None, annotations_directory_path=None, data_yaml_path=None, min_image_area_percentage=0.0, max_image_area_percentage=1.0, approximation_percentage=0.0)

Exports the dataset to YOLO format. This method saves the images and their corresponding annotations in YOLO format.

Parameters:

Name Type Description Default
images_directory_path Optional[str]

The path to the directory where the images should be saved. If not provided, images will not be saved.

None
annotations_directory_path Optional[str]

The path to the directory where the annotations in YOLO format should be saved. If not provided, annotations will not be saved.

None
data_yaml_path Optional[str]

The path where the data.yaml file should be saved. If not provided, the file will not be saved.

None
min_image_area_percentage float

The minimum percentage of detection area relative to the image area for a detection to be included. Argument is used only for segmentation datasets.

0.0
max_image_area_percentage float

The maximum percentage of detection area relative to the image area for a detection to be included. Argument is used only for segmentation datasets.

1.0
approximation_percentage float

The percentage of polygon points to be removed from the input polygon, in the range [0, 1). This is useful for simplifying the annotations. Argument is used only for segmentation datasets.

0.0
Source code in supervision/dataset/core.py
def as_yolo(
    self,
    images_directory_path: Optional[str] = None,
    annotations_directory_path: Optional[str] = None,
    data_yaml_path: Optional[str] = None,
    min_image_area_percentage: float = 0.0,
    max_image_area_percentage: float = 1.0,
    approximation_percentage: float = 0.0,
) -> None:
    """
    Exports the dataset to YOLO format. This method saves the
    images and their corresponding annotations in YOLO format.

    Args:
        images_directory_path (Optional[str]): The path to the
            directory where the images should be saved.
            If not provided, images will not be saved.
        annotations_directory_path (Optional[str]): The path to the
            directory where the annotations in
            YOLO format should be saved. If not provided,
            annotations will not be saved.
        data_yaml_path (Optional[str]): The path where the data.yaml
            file should be saved.
            If not provided, the file will not be saved.
        min_image_area_percentage (float): The minimum percentage of
            detection area relative to
            the image area for a detection to be included.
            Argument is used only for segmentation datasets.
        max_image_area_percentage (float): The maximum percentage
            of detection area relative to
            the image area for a detection to be included.
            Argument is used only for segmentation datasets.
        approximation_percentage (float): The percentage of polygon points to
            be removed from the input polygon, in the range [0, 1).
            This is useful for simplifying the annotations.
            Argument is used only for segmentation datasets.
    """
    if images_directory_path is not None:
        save_dataset_images(
            images_directory_path=images_directory_path, images=self.images
        )
    if annotations_directory_path is not None:
        save_yolo_annotations(
            annotations_directory_path=annotations_directory_path,
            images=self.images,
            annotations=self.annotations,
            min_image_area_percentage=min_image_area_percentage,
            max_image_area_percentage=max_image_area_percentage,
            approximation_percentage=approximation_percentage,
        )
    if data_yaml_path is not None:
        save_data_yaml(data_yaml_path=data_yaml_path, classes=self.classes)

from_coco(images_directory_path, annotations_path, force_masks=False) classmethod

Creates a Dataset instance from COCO formatted data.

Parameters:

Name Type Description Default
images_directory_path str

The path to the directory containing the images.

required
annotations_path str

The path to the json annotation files.

required
force_masks bool

If True, forces masks to be loaded for all annotations, regardless of whether they are present.

False

Returns:

Name Type Description
DetectionDataset DetectionDataset

A DetectionDataset instance containing the loaded images and annotations.

Examples:

import roboflow
from roboflow import Roboflow
import supervision as sv

roboflow.login()
rf = Roboflow()

project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
dataset = project.version(PROJECT_VERSION).download("coco")

ds = sv.DetectionDataset.from_coco(
    images_directory_path=f"{dataset.location}/train",
    annotations_path=f"{dataset.location}/train/_annotations.coco.json",
)

ds.classes
# ['dog', 'person']
Source code in supervision/dataset/core.py
@classmethod
def from_coco(
    cls,
    images_directory_path: str,
    annotations_path: str,
    force_masks: bool = False,
) -> DetectionDataset:
    """
    Creates a Dataset instance from COCO formatted data.

    Args:
        images_directory_path (str): The path to the
            directory containing the images.
        annotations_path (str): The path to the json annotation files.
        force_masks (bool, optional): If True,
            forces masks to be loaded for all annotations,
            regardless of whether they are present.

    Returns:
        DetectionDataset: A DetectionDataset instance containing
            the loaded images and annotations.

    Examples:
        ```python
        import roboflow
        from roboflow import Roboflow
        import supervision as sv

        roboflow.login()
        rf = Roboflow()

        project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
        dataset = project.version(PROJECT_VERSION).download("coco")

        ds = sv.DetectionDataset.from_coco(
            images_directory_path=f"{dataset.location}/train",
            annotations_path=f"{dataset.location}/train/_annotations.coco.json",
        )

        ds.classes
        # ['dog', 'person']
        ```
    """
    classes, images, annotations = load_coco_annotations(
        images_directory_path=images_directory_path,
        annotations_path=annotations_path,
        force_masks=force_masks,
    )
    return DetectionDataset(classes=classes, images=images, annotations=annotations)

from_pascal_voc(images_directory_path, annotations_directory_path, force_masks=False) classmethod

Creates a Dataset instance from PASCAL VOC formatted data.

Parameters:

Name Type Description Default
images_directory_path str

Path to the directory containing the images.

required
annotations_directory_path str

Path to the directory containing the PASCAL VOC XML annotations.

required
force_masks bool

If True, forces masks to be loaded for all annotations, regardless of whether they are present.

False

Returns:

Name Type Description
DetectionDataset DetectionDataset

A DetectionDataset instance containing the loaded images and annotations.

Examples:

import roboflow
from roboflow import Roboflow
import supervision as sv

roboflow.login()

rf = Roboflow()

project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
dataset = project.version(PROJECT_VERSION).download("voc")

ds = sv.DetectionDataset.from_pascal_voc(
    images_directory_path=f"{dataset.location}/train/images",
    annotations_directory_path=f"{dataset.location}/train/labels"
)

ds.classes
# ['dog', 'person']
Source code in supervision/dataset/core.py
@classmethod
def from_pascal_voc(
    cls,
    images_directory_path: str,
    annotations_directory_path: str,
    force_masks: bool = False,
) -> DetectionDataset:
    """
    Creates a Dataset instance from PASCAL VOC formatted data.

    Args:
        images_directory_path (str): Path to the directory containing the images.
        annotations_directory_path (str): Path to the directory
            containing the PASCAL VOC XML annotations.
        force_masks (bool, optional): If True, forces masks to
            be loaded for all annotations, regardless of whether they are present.

    Returns:
        DetectionDataset: A DetectionDataset instance containing
            the loaded images and annotations.

    Examples:
        ```python
        import roboflow
        from roboflow import Roboflow
        import supervision as sv

        roboflow.login()

        rf = Roboflow()

        project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
        dataset = project.version(PROJECT_VERSION).download("voc")

        ds = sv.DetectionDataset.from_pascal_voc(
            images_directory_path=f"{dataset.location}/train/images",
            annotations_directory_path=f"{dataset.location}/train/labels"
        )

        ds.classes
        # ['dog', 'person']
        ```
    """

    classes, images, annotations = load_pascal_voc_annotations(
        images_directory_path=images_directory_path,
        annotations_directory_path=annotations_directory_path,
        force_masks=force_masks,
    )

    return DetectionDataset(classes=classes, images=images, annotations=annotations)

from_yolo(images_directory_path, annotations_directory_path, data_yaml_path, force_masks=False) classmethod

Creates a Dataset instance from YOLO formatted data.

Parameters:

Name Type Description Default
images_directory_path str

The path to the directory containing the images.

required
annotations_directory_path str

The path to the directory containing the YOLO annotation files.

required
data_yaml_path str

The path to the data YAML file containing class information.

required
force_masks bool

If True, forces masks to be loaded for all annotations, regardless of whether they are present.

False

Returns:

Name Type Description
DetectionDataset DetectionDataset

A DetectionDataset instance containing the loaded images and annotations.

Examples:

import roboflow
from roboflow import Roboflow
import supervision as sv

roboflow.login()
rf = Roboflow()

project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
dataset = project.version(PROJECT_VERSION).download("yolov5")

ds = sv.DetectionDataset.from_yolo(
    images_directory_path=f"{dataset.location}/train/images",
    annotations_directory_path=f"{dataset.location}/train/labels",
    data_yaml_path=f"{dataset.location}/data.yaml"
)

ds.classes
# ['dog', 'person']
Source code in supervision/dataset/core.py
@classmethod
def from_yolo(
    cls,
    images_directory_path: str,
    annotations_directory_path: str,
    data_yaml_path: str,
    force_masks: bool = False,
) -> DetectionDataset:
    """
    Creates a Dataset instance from YOLO formatted data.

    Args:
        images_directory_path (str): The path to the
            directory containing the images.
        annotations_directory_path (str): The path to the directory
            containing the YOLO annotation files.
        data_yaml_path (str): The path to the data
            YAML file containing class information.
        force_masks (bool, optional): If True, forces
            masks to be loaded for all annotations,
            regardless of whether they are present.

    Returns:
        DetectionDataset: A DetectionDataset instance
            containing the loaded images and annotations.

    Examples:
        ```python
        import roboflow
        from roboflow import Roboflow
        import supervision as sv

        roboflow.login()
        rf = Roboflow()

        project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
        dataset = project.version(PROJECT_VERSION).download("yolov5")

        ds = sv.DetectionDataset.from_yolo(
            images_directory_path=f"{dataset.location}/train/images",
            annotations_directory_path=f"{dataset.location}/train/labels",
            data_yaml_path=f"{dataset.location}/data.yaml"
        )

        ds.classes
        # ['dog', 'person']
        ```
    """
    classes, images, annotations = load_yolo_annotations(
        images_directory_path=images_directory_path,
        annotations_directory_path=annotations_directory_path,
        data_yaml_path=data_yaml_path,
        force_masks=force_masks,
    )
    return DetectionDataset(classes=classes, images=images, annotations=annotations)

merge(dataset_list) classmethod

Merge a list of DetectionDataset objects into a single DetectionDataset object.

This method takes a list of DetectionDataset objects and combines their respective fields (classes, images, annotations) into a single DetectionDataset object.

Parameters:

Name Type Description Default
dataset_list List[DetectionDataset]

A list of DetectionDataset objects to merge.

required

Returns:

Type Description
DetectionDataset

A single DetectionDataset object containing

DetectionDataset

the merged data from the input list.

Examples:

import supervision as sv

ds_1 = sv.DetectionDataset(...)
len(ds_1)
# 100
ds_1.classes
# ['dog', 'person']

ds_2 = sv.DetectionDataset(...)
len(ds_2)
# 200
ds_2.classes
# ['cat']

ds_merged = sv.DetectionDataset.merge([ds_1, ds_2])
len(ds_merged)
# 300
ds_merged.classes
# ['cat', 'dog', 'person']
Source code in supervision/dataset/core.py
@classmethod
def merge(cls, dataset_list: List[DetectionDataset]) -> DetectionDataset:
    """
    Merge a list of `DetectionDataset` objects into a single
        `DetectionDataset` object.

    This method takes a list of `DetectionDataset` objects and combines
    their respective fields (`classes`, `images`,
    `annotations`) into a single `DetectionDataset` object.

    Args:
        dataset_list (List[DetectionDataset]): A list of `DetectionDataset`
            objects to merge.

    Returns:
        (DetectionDataset): A single `DetectionDataset` object containing
        the merged data from the input list.

    Examples:
        ```python
        import supervision as sv

        ds_1 = sv.DetectionDataset(...)
        len(ds_1)
        # 100
        ds_1.classes
        # ['dog', 'person']

        ds_2 = sv.DetectionDataset(...)
        len(ds_2)
        # 200
        ds_2.classes
        # ['cat']

        ds_merged = sv.DetectionDataset.merge([ds_1, ds_2])
        len(ds_merged)
        # 300
        ds_merged.classes
        # ['cat', 'dog', 'person']
        ```
    """
    merged_images, merged_annotations = {}, {}
    class_lists = [dataset.classes for dataset in dataset_list]
    merged_classes = merge_class_lists(class_lists=class_lists)

    for dataset in dataset_list:
        class_index_mapping = build_class_index_mapping(
            source_classes=dataset.classes, target_classes=merged_classes
        )
        for image_name, image, detections in dataset:
            if image_name in merged_annotations:
                raise ValueError(
                    f"Image name {image_name} is not unique across datasets."
                )

            merged_images[image_name] = image
            merged_annotations[image_name] = map_detections_class_id(
                source_to_target_mapping=class_index_mapping,
                detections=detections,
            )

    return cls(
        classes=merged_classes, images=merged_images, annotations=merged_annotations
    )

split(split_ratio=0.8, random_state=None, shuffle=True)

Splits the dataset into two parts (training and testing) using the provided split_ratio.

Parameters:

Name Type Description Default
split_ratio float

The ratio of the training set to the entire dataset.

0.8
random_state int

The seed for the random number generator. This is used for reproducibility.

None
shuffle bool

Whether to shuffle the data before splitting.

True

Returns:

Type Description
Tuple[DetectionDataset, DetectionDataset]

Tuple[DetectionDataset, DetectionDataset]: A tuple containing the training and testing datasets.

Examples:

import supervision as sv

ds = sv.DetectionDataset(...)
train_ds, test_ds = ds.split(split_ratio=0.7, random_state=42, shuffle=True)
len(train_ds), len(test_ds)
# (700, 300)
Source code in supervision/dataset/core.py
def split(
    self, split_ratio=0.8, random_state=None, shuffle: bool = True
) -> Tuple[DetectionDataset, DetectionDataset]:
    """
    Splits the dataset into two parts (training and testing)
        using the provided split_ratio.

    Args:
        split_ratio (float, optional): The ratio of the training
            set to the entire dataset.
        random_state (int, optional): The seed for the random number generator.
            This is used for reproducibility.
        shuffle (bool, optional): Whether to shuffle the data before splitting.

    Returns:
        Tuple[DetectionDataset, DetectionDataset]: A tuple containing
            the training and testing datasets.

    Examples:
        ```python
        import supervision as sv

        ds = sv.DetectionDataset(...)
        train_ds, test_ds = ds.split(split_ratio=0.7, random_state=42, shuffle=True)
        len(train_ds), len(test_ds)
        # (700, 300)
        ```
    """

    image_names = list(self.images.keys())
    train_names, test_names = train_test_split(
        data=image_names,
        train_ratio=split_ratio,
        random_state=random_state,
        shuffle=shuffle,
    )

    train_dataset = DetectionDataset(
        classes=self.classes,
        images={name: self.images[name] for name in train_names},
        annotations={name: self.annotations[name] for name in train_names},
    )
    test_dataset = DetectionDataset(
        classes=self.classes,
        images={name: self.images[name] for name in test_names},
        annotations={name: self.annotations[name] for name in test_names},
    )
    return train_dataset, test_dataset

ClassificationDataset

Bases: BaseDataset

Dataclass containing information about a classification dataset.

Attributes:

Name Type Description
classes List[str]

List containing dataset class names.

images Dict[str, ndarray]

Dictionary mapping image name to image.

annotations Dict[str, Detections]

Dictionary mapping image name to annotations.

Source code in supervision/dataset/core.py
@dataclass
class ClassificationDataset(BaseDataset):
    """
    Dataclass containing information about a classification dataset.

    Attributes:
        classes (List[str]): List containing dataset class names.
        images (Dict[str, np.ndarray]): Dictionary mapping image name to image.
        annotations (Dict[str, Detections]): Dictionary mapping
            image name to annotations.
    """

    classes: List[str]
    images: Dict[str, np.ndarray]
    annotations: Dict[str, Classifications]

    def __len__(self) -> int:
        return len(self.images)

    def split(
        self, split_ratio=0.8, random_state=None, shuffle: bool = True
    ) -> Tuple[ClassificationDataset, ClassificationDataset]:
        """
        Splits the dataset into two parts (training and testing)
            using the provided split_ratio.

        Args:
            split_ratio (float, optional): The ratio of the training
                set to the entire dataset.
            random_state (int, optional): The seed for the
                random number generator. This is used for reproducibility.
            shuffle (bool, optional): Whether to shuffle the data before splitting.

        Returns:
            Tuple[ClassificationDataset, ClassificationDataset]: A tuple containing
            the training and testing datasets.

        Examples:
            ```python
            import supervision as sv

            cd = sv.ClassificationDataset(...)
            train_cd,test_cd = cd.split(split_ratio=0.7, random_state=42,shuffle=True)
            len(train_cd), len(test_cd)
            # (700, 300)
            ```
        """
        image_names = list(self.images.keys())
        train_names, test_names = train_test_split(
            data=image_names,
            train_ratio=split_ratio,
            random_state=random_state,
            shuffle=shuffle,
        )

        train_dataset = ClassificationDataset(
            classes=self.classes,
            images={name: self.images[name] for name in train_names},
            annotations={name: self.annotations[name] for name in train_names},
        )
        test_dataset = ClassificationDataset(
            classes=self.classes,
            images={name: self.images[name] for name in test_names},
            annotations={name: self.annotations[name] for name in test_names},
        )
        return train_dataset, test_dataset

    def as_folder_structure(self, root_directory_path: str) -> None:
        """
        Saves the dataset as a multi-class folder structure.

        Args:
            root_directory_path (str): The path to the directory
                where the dataset will be saved.
        """
        os.makedirs(root_directory_path, exist_ok=True)

        for class_name in self.classes:
            os.makedirs(os.path.join(root_directory_path, class_name), exist_ok=True)

        for image_path in self.images:
            classification = self.annotations[image_path]
            image = self.images[image_path]
            image_name = Path(image_path).name
            class_id = (
                classification.class_id[0]
                if classification.confidence is None
                else classification.get_top_k(1)[0][0]
            )
            class_name = self.classes[class_id]
            image_path = os.path.join(root_directory_path, class_name, image_name)
            cv2.imwrite(image_path, image)

    @classmethod
    def from_folder_structure(cls, root_directory_path: str) -> ClassificationDataset:
        """
        Load data from a multiclass folder structure into a ClassificationDataset.

        Args:
            root_directory_path (str): The path to the dataset directory.

        Returns:
            ClassificationDataset: The dataset.

        Examples:
            ```python
            import roboflow
            from roboflow import Roboflow
            import supervision as sv

            roboflow.login()
            rf = Roboflow()

            project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
            dataset = project.version(PROJECT_VERSION).download("folder")

            cd = sv.ClassificationDataset.from_folder_structure(
                root_directory_path=f"{dataset.location}/train"
            )
            ```
        """
        classes = os.listdir(root_directory_path)
        classes = sorted(set(classes))

        images = {}
        annotations = {}

        for class_name in classes:
            class_id = classes.index(class_name)

            for image in os.listdir(os.path.join(root_directory_path, class_name)):
                image_path = str(os.path.join(root_directory_path, class_name, image))
                images[image_path] = cv2.imread(image_path)
                annotations[image_path] = Classifications(
                    class_id=np.array([class_id]),
                )

        return cls(
            classes=classes,
            images=images,
            annotations=annotations,
        )

Functions

as_folder_structure(root_directory_path)

Saves the dataset as a multi-class folder structure.

Parameters:

Name Type Description Default
root_directory_path str

The path to the directory where the dataset will be saved.

required
Source code in supervision/dataset/core.py
def as_folder_structure(self, root_directory_path: str) -> None:
    """
    Saves the dataset as a multi-class folder structure.

    Args:
        root_directory_path (str): The path to the directory
            where the dataset will be saved.
    """
    os.makedirs(root_directory_path, exist_ok=True)

    for class_name in self.classes:
        os.makedirs(os.path.join(root_directory_path, class_name), exist_ok=True)

    for image_path in self.images:
        classification = self.annotations[image_path]
        image = self.images[image_path]
        image_name = Path(image_path).name
        class_id = (
            classification.class_id[0]
            if classification.confidence is None
            else classification.get_top_k(1)[0][0]
        )
        class_name = self.classes[class_id]
        image_path = os.path.join(root_directory_path, class_name, image_name)
        cv2.imwrite(image_path, image)

from_folder_structure(root_directory_path) classmethod

Load data from a multiclass folder structure into a ClassificationDataset.

Parameters:

Name Type Description Default
root_directory_path str

The path to the dataset directory.

required

Returns:

Name Type Description
ClassificationDataset ClassificationDataset

The dataset.

Examples:

import roboflow
from roboflow import Roboflow
import supervision as sv

roboflow.login()
rf = Roboflow()

project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
dataset = project.version(PROJECT_VERSION).download("folder")

cd = sv.ClassificationDataset.from_folder_structure(
    root_directory_path=f"{dataset.location}/train"
)
Source code in supervision/dataset/core.py
@classmethod
def from_folder_structure(cls, root_directory_path: str) -> ClassificationDataset:
    """
    Load data from a multiclass folder structure into a ClassificationDataset.

    Args:
        root_directory_path (str): The path to the dataset directory.

    Returns:
        ClassificationDataset: The dataset.

    Examples:
        ```python
        import roboflow
        from roboflow import Roboflow
        import supervision as sv

        roboflow.login()
        rf = Roboflow()

        project = rf.workspace(WORKSPACE_ID).project(PROJECT_ID)
        dataset = project.version(PROJECT_VERSION).download("folder")

        cd = sv.ClassificationDataset.from_folder_structure(
            root_directory_path=f"{dataset.location}/train"
        )
        ```
    """
    classes = os.listdir(root_directory_path)
    classes = sorted(set(classes))

    images = {}
    annotations = {}

    for class_name in classes:
        class_id = classes.index(class_name)

        for image in os.listdir(os.path.join(root_directory_path, class_name)):
            image_path = str(os.path.join(root_directory_path, class_name, image))
            images[image_path] = cv2.imread(image_path)
            annotations[image_path] = Classifications(
                class_id=np.array([class_id]),
            )

    return cls(
        classes=classes,
        images=images,
        annotations=annotations,
    )

split(split_ratio=0.8, random_state=None, shuffle=True)

Splits the dataset into two parts (training and testing) using the provided split_ratio.

Parameters:

Name Type Description Default
split_ratio float

The ratio of the training set to the entire dataset.

0.8
random_state int

The seed for the random number generator. This is used for reproducibility.

None
shuffle bool

Whether to shuffle the data before splitting.

True

Returns:

Type Description
ClassificationDataset

Tuple[ClassificationDataset, ClassificationDataset]: A tuple containing

ClassificationDataset

the training and testing datasets.

Examples:

import supervision as sv

cd = sv.ClassificationDataset(...)
train_cd,test_cd = cd.split(split_ratio=0.7, random_state=42,shuffle=True)
len(train_cd), len(test_cd)
# (700, 300)
Source code in supervision/dataset/core.py
def split(
    self, split_ratio=0.8, random_state=None, shuffle: bool = True
) -> Tuple[ClassificationDataset, ClassificationDataset]:
    """
    Splits the dataset into two parts (training and testing)
        using the provided split_ratio.

    Args:
        split_ratio (float, optional): The ratio of the training
            set to the entire dataset.
        random_state (int, optional): The seed for the
            random number generator. This is used for reproducibility.
        shuffle (bool, optional): Whether to shuffle the data before splitting.

    Returns:
        Tuple[ClassificationDataset, ClassificationDataset]: A tuple containing
        the training and testing datasets.

    Examples:
        ```python
        import supervision as sv

        cd = sv.ClassificationDataset(...)
        train_cd,test_cd = cd.split(split_ratio=0.7, random_state=42,shuffle=True)
        len(train_cd), len(test_cd)
        # (700, 300)
        ```
    """
    image_names = list(self.images.keys())
    train_names, test_names = train_test_split(
        data=image_names,
        train_ratio=split_ratio,
        random_state=random_state,
        shuffle=shuffle,
    )

    train_dataset = ClassificationDataset(
        classes=self.classes,
        images={name: self.images[name] for name in train_names},
        annotations={name: self.annotations[name] for name in train_names},
    )
    test_dataset = ClassificationDataset(
        classes=self.classes,
        images={name: self.images[name] for name in test_names},
        annotations={name: self.annotations[name] for name in test_names},
    )
    return train_dataset, test_dataset

Comments