PointcloudDataset¶

class PointcloudDataset(directory, mode=None, parents=None, dataset_id=None, api=None)[source]¶

Bases: VideoDataset

A dataset directory for point cloud items inside a local Supervisely point cloud project.

Stores point clouds, their annotations and (optionally) related images for each item.

PointcloudDataset is a dataset for pointcloud data. PointcloudDataset object is immutable.

Parameters:

directory : str¶: Path to dataset directory.
mode=None¶: Determines working mode for the given dataset.
parents : List[str]¶: List of parent directories, e.g. [“ds1”, “ds2”, “ds3”].
dataset_id : Optional[int]¶: Dataset ID if the Dataset is opened in API mode. If dataset_id is specified then api must be specified as well.
api=None¶: API object if the Dataset is opened in API mode.

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/ptc_project/ds0"
ds = sly.PointcloudDataset(dataset_path, sly.OpenMode.READ)
print(ds.project_dir)

Methods

`add_item_file`	Adds given item file to dataset items directory, and adds given annotation to dataset annotations directory.
`add_item_file_async`	Adds given item file to dataset items directory, and adds given annotation to dataset annotations directory.
`add_item_np`	Adds given numpy array as a pointcloud to dataset items directory, and adds given annotation to dataset ann directory.
`add_item_raw_bytes`	Not available for PointcloudDataset class object.
`add_item_raw_bytes_async`	Adds given binary object as an image to dataset items directory, and adds given annotation to dataset ann directory.
`datasets_dir`	Returns the name of the datasets directory.
`delete_item`	Delete pointcloud, annotation, pointcloud info and related images from PointcloudDataset.
`generate_item_path`	Generates full path to the given item.
`get_ann`	Read pointcloud annotation of item from json.
`get_ann_path`	Path to the given annotation.
`get_blob_img_bytes`	Get image bytes from blob file.
`get_blob_img_np`	Get image as numpy array from blob file.
`get_classes_stats`
`get_image_info`	Not available for PointcloudDataset class object.
`get_img_info_path`	Not available for PointcloudDataset class object.
`get_img_path`	Not available for PointcloudDataset class object.
`get_item_info`	Information for Item with given name.
`get_item_info_path`	Get path to the item info json file without checking if the file exists.
`get_item_meta_path`	Get path to the item info json file without checking if the file exists.
`get_item_path`	Path to the given item.
`get_item_paths`
`get_items_names`	List of video dataset item names.
`get_pointcloud_info`	Information for Pointcloud with given name.
`get_pointcloud_info_path`
`get_pointcloud_path`	Path to the given pointcloud.
`get_related_image_figures`	Get figures from related image with given name.
`get_related_images`
`get_related_images_path`
`get_seg_path`	Not available for PointcloudDataset class object.
`get_url`	Get URL to dataset items list in Supervisely.
`ignorable_dirs`	Returns the list of ignorable directories.
`item_exists`	Checks if given item name belongs to the video dataset.
`items`	This method is used to iterate over dataset items, receiving item name, path to image and path to annotation json file.
`key`	Returns the key (name) of the dataset.
`set_ann`	Replaces given annotation for given item name to dataset annotations directory in json format.
`set_ann_async`	Replaces given annotation for given item name to dataset annotations directory in json format.
`set_ann_dict`	Replaces given annotation json for given item name to dataset annotations directory in json format.
`set_ann_dict_async`	Replaces given annotation json for given item name to dataset annotations directory in json format.
`set_ann_file`	Replaces given annotation json file for given item name to dataset annotations directory in json format.
`set_ann_file_async`	Replaces given annotation json file for given item name to dataset annotations directory in json format.
`to_coco`	Convert Supervisely dataset to COCO format.
`to_pascal_voc`	Convert Supervisely dataset to Pascal VOC format.
`to_yolo`	Convert Supervisely dataset to YOLO format.

Attributes

`ann_dir`	Path to the video dataset annotations directory.
`ann_dir_name`	str: Annotations directory name
`blob_dir_name`
`blob_offsets`	List of paths to the dataset blob offset files.
`datasets_dir_name`
`directory`	Path to the video dataset directory.
`image_infos`	If the dataset is opened from the API, returns the list of ImageInfo objects.
`img_dir`	Not available for PointcloudDataset class object.
`img_info_dir`	Not available for PointcloudDataset class object.
`item_dir`	Path to the video dataset items directory.
`item_dir_name`	str: Items data directory name
`item_info_dir`	Path to the video dataset item with items info.
`item_info_dir_name`	str: Items info directory name
`meta_dir`	Path to the dataset segmentation masks directory.
`meta_dir_name`
`metadata_dir_name`	str: Metadata directory name
`metadata_directory`	Path to the video dataset metadata directory.
`name`	Video Dataset name.
`path`	Returns a relative local path to the dataset.
`pointcloud_dir`
`pointcloud_info_dir`
`project_dir`	Path to the video project containing the video dataset.
`related_images_dir_name`	str: Related images directory name
`seg_dir`	Not available for PointcloudDataset class object.
`seg_dir_name`	str: Segmentation masks directory name
`short_name`	Short dataset name, which does not include it's parents.

annotation_class¶: alias of PointcloudAnnotation

item_info_class¶: alias of VideoInfo

item_info_type¶: alias of PointcloudInfo

classmethod datasets_dir()¶: Returns the name of the datasets directory.

classmethod ignorable_dirs()¶: Returns the list of ignorable directories.

static get_url(project_id, dataset_id)¶

Get URL to dataset items list in Supervisely.

Parameters:

project_id : int¶: VideoProject ID in Supervisely.
dataset_id : int¶: VideoDataset ID in Supervisely.

Returns:

URL to dataset items list.

Return type:

str

Usage Example:

from supervisely import VideoDataset

project_id = 10093
dataset_id = 45330
ds_items_link = VideoDataset.get_url(project_id, dataset_id)

print(ds_items_link)
# Output: "/projects/10093/datasets/45330"

add_item_file(item_name, item_path, ann=None, _validate_item=True, _use_hardlink=False, item_info=None)[source]¶

Adds given item file to dataset items directory, and adds given annotation to dataset annotations directory. if ann is None, creates empty annotation file.

Parameters:

item_name : str¶: Item name.
item_path : str¶: Path to the item.
ann=None¶: Pointcloud annotation object or path to annotation json file.
_validate_item : bool, optional¶: Checks input files format.
_use_hardlink : bool, optional¶: If True creates a hardlink pointing to src named dst, otherwise don’t.
item_info=None¶: Pointcloud info object or Pointcloud info object converted to dict or path to item info json file for copying to dataset item info directory.

Returns:

None

Return type:

None

Raises:

RuntimeError – if item_name already exists in dataset or item name has unsupported extension.

async add_item_file_async(item_name, item_path, ann=None, _validate_item=True, _use_hardlink=False, item_info=None, img_info=None)¶

Adds given item file to dataset items directory, and adds given annotation to dataset annotations directory. If ann is None, creates empty annotation file.

Parameters:

item_name : str¶: Item name.
item_path : str¶: Path to the item.
ann=None¶: Annotation object or path to annotation json file.
_validate_item : bool, optional¶: Checks input files format.
_use_hardlink : bool, optional¶: If True creates a hardlink pointing to src named dst, otherwise don’t.
item_info=None¶: ImageInfo object or ImageInfo object converted to dict or path to item info json file for copying to dataset item info directory.
img_info=None¶: Deprecated version of item_info parameter. Can be removed in future versions.

Returns:

None

Return type:

NoneType

Raises:

RuntimeError – if item_name already exists in dataset or item name has unsupported extension.

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
ds = sly.Dataset(dataset_path, sly.OpenMode.READ)

ann = "/home/admin/work/supervisely/projects/lemons_annotated/ds1/ann/IMG_8888.jpeg.json"
loop = sly.utils.get_or_create_event_loop()
loop.run_until_complete(
    ds.add_item_file_async("IMG_8888.jpeg", "/home/admin/work/supervisely/projects/lemons_annotated/ds1/img/IMG_8888.jpeg", ann=ann)
)
print(ds.item_exists("IMG_8888.jpeg"))
# Output: True

add_item_np(item_name, pointcloud, ann=None, item_info=None)[source]¶

Adds given numpy array as a pointcloud to dataset items directory, and adds given annotation to dataset ann directory. if ann is None, creates empty annotation file.

Parameters:

item_name : str¶: Item name.
pointcloud : np.ndarray¶: numpy Pointcloud array [N, 3], in (X, Y, Z) format.
ann=None¶: Pointcloud annotation object or path to annotation .json file.
item_info : NamedTuple, optional¶: NamedTuple containing information about pointcloud item.

Returns:

None

Return type:

None

Raises:

Exception – if item_name already exists in dataset or item name has unsupported extension

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/ptc_project/ds0"
ds = sly.PointcloudDataset(dataset_path, sly.OpenMode.READ)

pointcloud_path = "/home/admin/Pointclouds/ptc0.pcd"
img_np = sly.image.read(img_path)
ds.add_item_np("IMG_050.jpeg", img_np)

add_item_raw_bytes(item_name, item_raw_bytes, ann=None, img_info=None)[source]¶: Not available for PointcloudDataset class object. :raises NotImplementedError: in all cases.

async add_item_raw_bytes_async(item_name, item_raw_bytes, ann=None, img_info=None)¶

Adds given binary object as an image to dataset items directory, and adds given annotation to dataset ann directory. If ann is None, creates empty annotation file.

Parameters:

item_name : str¶: Item name.
item_raw_bytes : bytes¶: Binary object.
ann=None¶: Annotation object or path to annotation json file.
img_info=None¶: ImageInfo object or ImageInfo object converted to dict or path to item info json file for copying to dataset item info directory.

Returns:

None

Return type:

NoneType

Raises:

RuntimeError – if item_name already exists in dataset or item name has unsupported extension

Usage Example:

import supervisely as sly
from supervisely._utils import run_coroutine

dataset_path = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
ds = sly.Dataset(dataset_path, sly.OpenMode.READ)

img_path = "/home/admin/Pictures/Clouds.jpeg"
img_np = sly.image.read(img_path)
img_bytes = sly.image.write_bytes(img_np, "jpeg")
coroutine = ds.add_item_raw_bytes_async("IMG_050.jpeg", img_bytes)
run_coroutine(coroutine)

print(ds.item_exists("IMG_050.jpeg"))
# Output: True

delete_item(item_name)[source]¶

Delete pointcloud, annotation, pointcloud info and related images from PointcloudDataset.

Parameters:

item_name : str¶: Item name.

Returns:

True if successful, otherwise False

Return type:

bool

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/ptc_project/ds0"
ds = sly.PointcloudDataset(dataset_path, sly.OpenMode.READ)
result = dataset.delete_item("PTC_0748.pcd")
# Output: True

generate_item_path(item_name)¶

Generates full path to the given item.

Parameters:

item_name : str¶: Item name.

Returns:

Full path to the given item

Return type:

str

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
ds = sly.Dataset(dataset_path, sly.OpenMode.READ)

print(ds.generate_item_path("IMG_0748.jpeg"))
# Output: '/home/admin/work/supervisely/projects/lemons_annotated/ds1/img/IMG_0748.jpeg'

get_ann(item_name, project_meta, key_id_map=None)[source]¶

Read pointcloud annotation of item from json.

Parameters:

item_name : str¶: Pointcloud name.
project_meta¶: Project Meta.
key_id_map=None¶: Key ID map.

Returns:

Pointcloud annotation object

Return type:

PointcloudAnnotation

Raises:

RuntimeError – if item not found in the project

Usage Example:

import supervisely as sly

project_path = "/home/admin/work/supervisely/projects/pointcloud_project"
project = sly.PointcloudProject(project_path, sly.OpenMode.READ)

ds = project.datasets.get('ds1')

annotation = ds.get_ann("PTC_0056")
# Output: RuntimeError: Item PTC_0056 not found in the project.

annotation = ds.get_ann("PTC_0056.pcd")
print(type(annotation).__name__)
# Output: PointcloudAnnotation

get_ann_path(item_name)[source]¶

Path to the given annotation.

Parameters:

item_name : str¶: PointcloudAnnotation name.

Returns:

Path to the given annotation

Return type:

str

Raises:

RuntimeError – if item not found in the project

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/ptc_project/ds0"
ds = sly.PointcloudDataset(dataset_path, sly.OpenMode.READ)

ds.get_ann_path("PTC_0748")
# Output: RuntimeError: Item PTC_0748 not found in the project.

ds.get_ann_path("PTC_0748.pcd")
# Output: '/home/admin/work/supervisely/projects/ptc_project/ds0/ann/IMG_0748.pcd.json'

get_blob_img_bytes(image_name)¶

Get image bytes from blob file.

Parameters:

image_name : str¶: Image name with extension.

Returns:

Bytes of the image.

Return type:

bytes

Usage Example:

import supervisely as sly

dataset_path = "/path/to/project/lemons_annotated/ds1"
dataset = sly.Dataset(dataset_path, sly.OpenMode.READ)
image_name = "IMG_0748.jpeg"

img_bytes = dataset.get_blob_img_bytes(image_name)

get_blob_img_np(image_name)¶

Get image as numpy array from blob file.

Parameters:

image_name : str¶: Image name with extension.

Returns:

Numpy array of the image.

Return type:

numpy.ndarray

Usage Example:

import supervisely as sly

dataset_path = "/path/to/project/lemons_annotated/ds1"
dataset = sly.Dataset(dataset_path, sly.OpenMode.READ)
image_name = "IMG_0748.jpeg"
img_np = dataset.get_blob_img_np(image_name)

get_image_info(item_name)[source]¶: Not available for PointcloudDataset class object. :raises NotImplementedError: in all cases.

get_img_info_path(img_name)[source]¶: Not available for PointcloudDataset class object. :raises NotImplementedError: in all cases.

get_img_path(item_name)[source]¶: Not available for PointcloudDataset class object. :raises NotImplementedError: in all cases.

get_item_info(item_name)¶

Information for Item with given name.

Parameters:

item_name : str¶: Item name.

Returns:

VideoInfo object.

Return type:

VideoInfo

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)

print(ds.get_item_info("video_0748.mp4"))
# Output:
# VideoInfo(
#     id=198702499,
#     name='video_0748.mp4',
#     hash='ehYHLNFWmMNuF2fPUgnC/g/tkIIEjNIOhdbNLQXkE8Y=',
#     team_id=16087,
#     workspace_id=23821,
#     project_id=124974,
#     dataset_id=466639,
#     path_original='/h5un6l2bnaz1vj8a9qgms4-public/videos/w/7/i4/GZYoCs...9F3kyVJ7.mp4',
#     frames_to_timecodes=[0, 0.033367, 0.066733, 0.1001,...,10.777433, 10.8108, 10.844167],
#     frames_count=326,
#     frame_width=3840,
#     frame_height=2160,
#     created_at='2021-03-23T13:14:25.536Z',
#     updated_at='2021-03-23T13:16:43.300Z'
# )

get_item_info_path(item_name)¶

Get path to the item info json file without checking if the file exists.

Parameters:

item_name : str¶: Item name.

Returns:

Path to the given item info json file.

Return type:

str

Raises:

RuntimeError – if item not found in the project.

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)

print(ds.get_item_info_path("video_0748"))
# Output: RuntimeError: Item video_0748 not found in the project.

print(ds.get_item_info_path("video_0748.mp4"))
# Output: '/home/admin/work/supervisely/projects/videos_example/ds0/video_info/video_0748.mp4.json'

get_item_meta_path(item_name)¶

Get path to the item info json file without checking if the file exists.

Parameters:

item_name : str¶: Item name.

Returns:

Path to the given item info json file.

Return type:

str

Raises:

RuntimeError – if item not found in the project.

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
ds = sly.Dataset(dataset_path, sly.OpenMode.READ)

print(ds.get_item_info_path("IMG_0748"))
# Output: RuntimeError: Item IMG_0748 not found in the project.

print(ds.get_item_info_path("IMG_0748.jpeg"))
# Output: '/home/admin/work/supervisely/projects/lemons_annotated/ds1/img_info/IMG_0748.jpeg.json'

get_item_path(item_name)¶

Path to the given item.

Parameters:

item_name : str¶: Item name.

Returns:

Path to the given item.

Return type:

str

Raises:

RuntimeError – if item not found in the project

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)

print(ds.get_item_path("video_0748"))
# Output: RuntimeError: Item video_0748 not found in the project.

print(ds.get_item_path("video_0748.mp4"))
# Output: '/home/admin/work/supervisely/projects/videos_example/ds0/video/video_0748.mp4'

get_items_names()¶

List of video dataset item names.

Returns:

List of item names.

Return type:

List[str]

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)

print(ds.get_item_names())
# Output: ['video_0002.mp4', 'video_0005.mp4', 'video_0008.mp4', ...]

get_pointcloud_info(item_name)[source]¶

Information for Pointcloud with given name.

Parameters:

item_name : str¶: Pointcloud name.

Returns:

Pointcloud with information for the given Dataset

Return type:

NamedTuple

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/ptc_project/ds0"
ds = sly.PointcloudDataset(dataset_path, sly.OpenMode.READ)
info = ds.get_pointcloud_info("IMG_0748.pcd")

get_pointcloud_path(item_name)[source]¶

Path to the given pointcloud.

Parameters:

item_name : str¶: Pointcloud name

Returns:

Path to the given pointcloud

Return type:

str

Raises:

RuntimeError – if item not found in the project

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/ptc_project/ds0"
ds = sly.PointcloudDataset(dataset_path, sly.OpenMode.READ)

ds.get_pointcloud_path("PTC_0748")
# Output: RuntimeError: Item PTC_0748 not found in the project.

ds.get_pointcloud_path("PTC_0748.pcd")
# Output: '/home/admin/work/supervisely/projects/ptc_project/ds0/pointcloud/PTC_0748.pcd'

get_related_image_figures(item_name, img_name)[source]¶

Get figures from related image with given name.

Parameters:

item_name : str¶: Name of the item in the dataset.
img_name : str¶: Related image name.

Returns:

List of figures from related image.

Return type:

List[Dict]

Raises:

RuntimeError – if item not found in the project

get_seg_path(item_name)[source]¶: Not available for PointcloudDataset class object. :raises NotImplementedError: in all cases.

item_exists(item_name)¶

Checks if given item name belongs to the video dataset.

Parameters:

item_name : str¶: Item name.

Returns:

True if item exist, otherwise False.

Return type:

bool

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)

ds.item_exists("video_0748")     # False
ds.item_exists("video_0748.mp4") # True

items()¶

This method is used to iterate over dataset items, receiving item name, path to image and path to annotation json file. It is useful when you need to iterate over dataset items and get paths to images and annotations.

Returns:

Generator object, that yields tuple of item name, path to image and path to annotation json file.

Return type:

Generator[Tuple[str]]

Usage Example:

import supervisely as sly

input = "path/to/local/directory"
# Creating Supervisely project from local directory.
project = sly.Project(input, sly.OpenMode.READ)

for dataset in project.datasets:
    for item_name, image_path, ann_path in dataset.items():
        print(f"Item '{item_name}': image='{image_path}', ann='{ann_path}'")

key()¶: Returns the key (name) of the dataset.

set_ann(item_name, ann, key_id_map=None)¶

Replaces given annotation for given item name to dataset annotations directory in json format.

Parameters:

item_name : str¶: Item name.
ann¶: VideoAnnotation object.
key_id_map=None¶: KeyIdMap object.

Returns:

None

Return type:

NoneType

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)

height, width = 500, 700
new_ann = sly.VideoAnnotation((height, width), frames_count=0)
ds.set_ann("video_0748.mp4", new_ann)

async set_ann_async(item_name, ann)¶

Replaces given annotation for given item name to dataset annotations directory in json format.

Parameters:

item_name : str¶: Item name.
ann¶: Annotation object.

Returns:

None

Return type:

NoneType

Usage Example:

import supervisely as sly
from supervisely._utils import run_coroutine

dataset_path = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
ds = sly.Dataset(dataset_path, sly.OpenMode.READ)

height, width = 500, 700
new_ann = sly.Annotation((height, width))

coroutine = ds.set_ann_async("IMG_0748.jpeg", new_ann)
run_coroutine(coroutine)

set_ann_dict(item_name, ann)¶

Replaces given annotation json for given item name to dataset annotations directory in json format.

Parameters:

item_name : str¶: Item name.
ann : dict¶: Annotation as a dict in json format.

Returns:

None

Return type:

NoneType

Raises:

RuntimeError – if ann_path is not str

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
ds = sly.Dataset(dataset_path, sly.OpenMode.READ)

new_ann_json = {
    "description":"",
    "size":{"height":500, "width":700},
    "tags":[],
    "objects":[],
    "customBigData":{}
}

ds.set_ann_dict("IMG_8888.jpeg", new_ann_json)

async set_ann_dict_async(item_name, ann)¶

Replaces given annotation json for given item name to dataset annotations directory in json format.

Parameters:

item_name : str¶: Item name.
ann : dict¶: Annotation as a dict in json format.

Returns:

None

Return type:

NoneType

Raises:

RuntimeError – if ann_path is not str

Usage Example:

import supervisely as sly
from supervisely._utils import run_coroutine

dataset_path = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
ds = sly.Dataset(dataset_path, sly.OpenMode.READ)

new_ann_json = {
    "description":"",
    "size":{"height":500, "width":700},
    "tags":[],
    "objects":[],
    "customBigData":{}
}

coroutine = ds.set_ann_dict_async("IMG_8888.jpeg", new_ann_json)
run_coroutine(coroutine)

set_ann_file(item_name, ann_path)¶

Replaces given annotation json file for given item name to dataset annotations directory in json format.

Parameters:

item_name : str¶: Item Name.
ann_path : str¶: Path to the Annotation json file.

Returns:

None

Return type:

NoneType

Raises:

RuntimeError – if ann_path is not str

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
ds = sly.Dataset(dataset_path, sly.OpenMode.READ)

new_ann = "/home/admin/work/supervisely/projects/kiwi_annotated/ds1/ann/IMG_1812.jpeg.json"
ds.set_ann_file("IMG_1812.jpeg", new_ann)

async set_ann_file_async(item_name, ann_path)¶

Replaces given annotation json file for given item name to dataset annotations directory in json format.

Parameters:

item_name : str¶: Item Name.
ann_path : str¶: Path to the Annotation json file.

Returns:

None

Return type:

NoneType

Raises:

RuntimeError – if ann_path is not str

Usage Example:

import supervisely as sly
from supervisely._utils import run_coroutine

dataset_path = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
ds = sly.Dataset(dataset_path, sly.OpenMode.READ)
new_ann = "/home/admin/work/supervisely/projects/kiwi_annotated/ds1/ann/IMG_1812.jpeg.json"

coroutine = ds.set_ann_file_async("IMG_1812.jpeg", new_ann)
run_coroutine(coroutine)

to_coco(meta, return_type='path', dest_dir=None, copy_images=False, with_captions=False, log_progress=False, progress_cb=None)¶

Convert Supervisely dataset to COCO format.

Note: Depending on the return_type and with_captions parameters, the function returns different values.: If return_type is “path”, the COCO annotation files will be saved to the disk. If return_type is “dict”, the function returns COCO dataset in dictionary format. If with_captions is True, the function returns Tuple (instances and captions).

Parameters:

meta¶: Project meta information.
return_type : str, optional¶: Return type (path or dict).
dest_dir : str, optional¶: Path to save COCO dataset.
copy_images : bool, optional¶: If True, copies images to the COCO dataset directory.
with_captions : bool, optional¶: If True, returns captions
log_progress : str, optional¶: If True, log progress.
progress_cb=None¶: Progress callback.

Returns:

COCO dataset in dictionary format.

Return type:

dict

Usage Example:

import supervisely as sly

project_path = "/home/admin/work/supervisely/projects/lemons_annotated"
project = sly.Project(project_path, sly.OpenMode.READ)

for ds in project.datasets:
    dest_dir = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
    coco: Tuple[Dict, Dict] = ds.to_coco(project.meta, dest_dir=dest_dir)

to_pascal_voc(meta, dest_dir=None, train_val_split_coef=0.8, log_progress=False, progress_cb=None)¶

Convert Supervisely dataset to Pascal VOC format.

Parameters:

meta¶: Project meta information.
dest_dir : str, optional¶: Destination directory.
train_val_split_coef : float, optional¶: Coefficient for splitting images into train and validation sets.
log_progress : str, optional¶: If True, log progress.
progress_cb=None¶: Progress callback.

Returns:

None

Return type:

NoneType

Usage Example:

import supervisely as sly

project_path = "/home/admin/work/supervisely/projects/lemons_annotated"
project = sly.Project(project_path, sly.OpenMode.READ)

for ds in project.datasets:
    dest_dir = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
    ds.to_pascal_voc(project.meta, dest_dir=dest_dir)

to_yolo(meta, dest_dir=None, task_type='detect', log_progress=False, progress_cb=None, is_val=None)¶

Convert Supervisely dataset to YOLO format.

Parameters:

meta¶: Project meta information.
dest_dir : str, optional¶: Path to save YOLO dataset.
task_type : str, optional¶: Task type.
log_progress : str, optional¶: If True, log progress.
progress_cb=None¶: Progress callback.
is_val : bool, optional¶: If True, the dataset is a validation dataset.

Returns:

YOLO dataset in dictionary format.

Return type:

dict

Usage Example:

import supervisely as sly

project_path = "/home/admin/work/supervisely/projects/lemons_annotated"
project = sly.Project(project_path, sly.OpenMode.READ)

for ds in project.datasets:
    dest_dir = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
    ds.to_yolo(project.meta, dest_dir=dest_dir)

property ann_dir : str¶

Path to the video dataset annotations directory.

Returns:

Path to the video dataset directory with annotations.

Return type:

str

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)

print(ds.ann_dir)
# Output: '/home/admin/work/supervisely/projects/videos_example/ds0/ann'

ann_dir_name = 'ann'¶: str: Annotations directory name

property blob_offsets¶

List of paths to the dataset blob offset files.

Returns:: List of paths to the dataset blob offset files.
Return type:: List[str]

property directory : str¶

Path to the video dataset directory.

Returns:

Path to the video dataset directory.

Return type:

str

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)

print(ds.directory)
# Output: '/home/admin/work/supervisely/projects/videos_example/ds0'

property image_infos : list[supervisely.api.image_api.ImageInfo]¶

If the dataset is opened from the API, returns the list of ImageInfo objects. Otherwise raises an exception.

Raises:: ValueError – If the dataset is opened in local mode.
Returns:: List of ImageInfo objects.
Return type:: List[ImageInfo]

property img_dir : str¶: Not available for PointcloudDataset class object. :raises NotImplementedError: in all cases.

property img_info_dir¶: Not available for PointcloudDataset class object. :raises NotImplementedError: in all cases.

property item_dir : str¶

Path to the video dataset items directory.

Returns:

Path to the video dataset items directory.

Return type:

str

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)

print(ds.item_dir)
# Output: '/home/admin/work/supervisely/projects/videos_example/ds0/video'

item_dir_name = 'pointcloud'¶: str: Items data directory name

property item_info_dir¶

Path to the video dataset item with items info.

Returns:

Path to the video dataset directory with items info.

Return type:

str

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)

print(ds.item_info_dir)
# Output: '/home/admin/work/supervisely/projects/videos_example/ds0/video_info'

item_info_dir_name = 'pointcloud_info'¶: str: Items info directory name

property meta_dir¶

Path to the dataset segmentation masks directory.

Returns:

Path to the dataset directory with masks.

Return type:

str

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
ds = sly.Dataset(dataset_path, sly.OpenMode.READ)

print(ds.meta_dir)
# Output: '/home/admin/work/supervisely/projects/lemons_annotated/ds1/meta'

property metadata_directory : str¶

Path to the video dataset metadata directory.

Returns:

Path to the video dataset metadata directory.

Return type:

str

Usage example:

import supervisely as sly
dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)

print(ds.metadata_directory)
# Output: '/home/admin/work/supervisely/projects/videos_example/ds0/metadata'

property name : str¶

Video Dataset name.

Returns:

Video Dataset Name.

Return type:

str

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)
print(ds.name)
# Output: "ds0"

property path : str¶

Returns a relative local path to the dataset.

Returns:: Relative local path to the dataset.
Return type:: str

property project_dir : str¶

Path to the video project containing the video dataset.

Returns:

Path to the video project.

Return type:

str

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/videos_example/ds0"
ds = sly.VideoDataset(dataset_path, sly.OpenMode.READ)
print(ds.project_dir)
# Output: "/home/admin/work/supervisely/projects/videos_example"

related_images_dir_name = 'related_images'¶: str: Related images directory name

property seg_dir¶: Not available for PointcloudDataset class object. :raises NotImplementedError: in all cases.

seg_dir_name = None¶: str: Segmentation masks directory name

property short_name : str¶

Short dataset name, which does not include it’s parents. To get the full name of the dataset, use name.

Returns:

Dataset Name.

Return type:

str

Usage Example:

import supervisely as sly

dataset_path = "/home/admin/work/supervisely/projects/lemons_annotated/ds1"
ds = sly.Dataset(dataset_path, sly.OpenMode.READ)
print(ds.name)
# Output: "ds1"