diff --git a/torchvision/datasets/_optical_flow.py b/torchvision/datasets/_optical_flow.py
index c7663258899..40d25583942 100644
--- a/torchvision/datasets/_optical_flow.py
+++ b/torchvision/datasets/_optical_flow.py
@@ -13,7 +13,6 @@
 from .utils import _read_pfm, verify_str_arg
 from .vision import VisionDataset
 
-
 T1 = Tuple[Image.Image, Image.Image, Optional[np.ndarray], Optional[np.ndarray]]
 T2 = Tuple[Image.Image, Image.Image, Optional[np.ndarray]]
 
@@ -33,7 +32,7 @@ class FlowDataset(ABC, VisionDataset):
     # and it's up to whatever consumes the dataset to decide what valid_flow_mask should be.
     _has_builtin_flow_mask = False
 
-    def __init__(self, root: str, transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], transforms: Optional[Callable] = None) -> None:
 
         super().__init__(root=root)
         self.transforms = transforms
@@ -113,7 +112,7 @@ class Sintel(FlowDataset):
                         ...
 
     Args:
-        root (string): Root directory of the Sintel Dataset.
+        root (str or ``pathlib.Path``): Root directory of the Sintel Dataset.
         split (string, optional): The dataset split, either "train" (default) or "test"
         pass_name (string, optional): The pass to use, either "clean" (default), "final", or "both". See link above for
             details on the different passes.
@@ -125,7 +124,7 @@ class Sintel(FlowDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         pass_name: str = "clean",
         transforms: Optional[Callable] = None,
@@ -183,7 +182,7 @@ class KittiFlow(FlowDataset):
                     flow_occ
 
     Args:
-        root (string): Root directory of the KittiFlow Dataset.
+        root (str or ``pathlib.Path``): Root directory of the KittiFlow Dataset.
         split (string, optional): The dataset split, either "train" (default) or "test"
         transforms (callable, optional): A function/transform that takes in
             ``img1, img2, flow, valid_flow_mask`` and returns a transformed version.
@@ -191,7 +190,7 @@ class KittiFlow(FlowDataset):
 
     _has_builtin_flow_mask = True
 
-    def __init__(self, root: str, split: str = "train", transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], split: str = "train", transforms: Optional[Callable] = None) -> None:
         super().__init__(root=root, transforms=transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "test"))
@@ -248,7 +247,7 @@ class FlyingChairs(FlowDataset):
 
 
     Args:
-        root (string): Root directory of the FlyingChairs Dataset.
+        root (str or ``pathlib.Path``): Root directory of the FlyingChairs Dataset.
         split (string, optional): The dataset split, either "train" (default) or "val"
         transforms (callable, optional): A function/transform that takes in
             ``img1, img2, flow, valid_flow_mask`` and returns a transformed version.
@@ -256,7 +255,7 @@ class FlyingChairs(FlowDataset):
             return a built-in valid mask, such as :class:`~torchvision.datasets.KittiFlow`.
     """
 
-    def __init__(self, root: str, split: str = "train", transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], split: str = "train", transforms: Optional[Callable] = None) -> None:
         super().__init__(root=root, transforms=transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "val"))
@@ -316,7 +315,7 @@ class FlyingThings3D(FlowDataset):
                     TRAIN
 
     Args:
-        root (string): Root directory of the intel FlyingThings3D Dataset.
+        root (str or ``pathlib.Path``): Root directory of the intel FlyingThings3D Dataset.
         split (string, optional): The dataset split, either "train" (default) or "test"
         pass_name (string, optional): The pass to use, either "clean" (default) or "final" or "both". See link above for
             details on the different passes.
@@ -329,7 +328,7 @@ class FlyingThings3D(FlowDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         pass_name: str = "clean",
         camera: str = "left",
@@ -411,7 +410,7 @@ class HD1K(FlowDataset):
                     image_2
 
     Args:
-        root (string): Root directory of the HD1K Dataset.
+        root (str or ``pathlib.Path``): Root directory of the HD1K Dataset.
         split (string, optional): The dataset split, either "train" (default) or "test"
         transforms (callable, optional): A function/transform that takes in
             ``img1, img2, flow, valid_flow_mask`` and returns a transformed version.
@@ -419,7 +418,7 @@ class HD1K(FlowDataset):
 
     _has_builtin_flow_mask = True
 
-    def __init__(self, root: str, split: str = "train", transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], split: str = "train", transforms: Optional[Callable] = None) -> None:
         super().__init__(root=root, transforms=transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "test"))
diff --git a/torchvision/datasets/_stereo_matching.py b/torchvision/datasets/_stereo_matching.py
index c180e2e1eb8..6a3f563a2da 100644
--- a/torchvision/datasets/_stereo_matching.py
+++ b/torchvision/datasets/_stereo_matching.py
@@ -27,7 +27,7 @@ class StereoMatchingDataset(ABC, VisionDataset):
 
     _has_built_in_disparity_mask = False
 
-    def __init__(self, root: str, transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], transforms: Optional[Callable] = None) -> None:
         """
         Args:
             root(str): Root directory of the dataset.
@@ -159,11 +159,11 @@ class CarlaStereo(StereoMatchingDataset):
                     ...
 
     Args:
-        root (string): Root directory where `carla-highres` is located.
+        root (str or ``pathlib.Path``): Root directory where `carla-highres` is located.
         transforms (callable, optional): A function/transform that takes in a sample and returns a transformed version.
     """
 
-    def __init__(self, root: str, transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], transforms: Optional[Callable] = None) -> None:
         super().__init__(root, transforms)
 
         root = Path(root) / "carla-highres"
@@ -233,14 +233,14 @@ class Kitti2012Stereo(StereoMatchingDataset):
                     calib
 
     Args:
-        root (string): Root directory where `Kitti2012` is located.
+        root (str or ``pathlib.Path``): Root directory where `Kitti2012` is located.
         split (string, optional): The dataset split of scenes, either "train" (default) or "test".
         transforms (callable, optional): A function/transform that takes in a sample and returns a transformed version.
     """
 
     _has_built_in_disparity_mask = True
 
-    def __init__(self, root: str, split: str = "train", transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], split: str = "train", transforms: Optional[Callable] = None) -> None:
         super().__init__(root, transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "test"))
@@ -321,14 +321,14 @@ class Kitti2015Stereo(StereoMatchingDataset):
                     calib
 
     Args:
-        root (string): Root directory where `Kitti2015` is located.
+        root (str or ``pathlib.Path``): Root directory where `Kitti2015` is located.
         split (string, optional): The dataset split of scenes, either "train" (default) or "test".
         transforms (callable, optional): A function/transform that takes in a sample and returns a transformed version.
     """
 
     _has_built_in_disparity_mask = True
 
-    def __init__(self, root: str, split: str = "train", transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], split: str = "train", transforms: Optional[Callable] = None) -> None:
         super().__init__(root, transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "test"))
@@ -420,7 +420,7 @@ class Middlebury2014Stereo(StereoMatchingDataset):
                     ...
 
     Args:
-        root (string): Root directory of the Middleburry 2014 Dataset.
+        root (str or ``pathlib.Path``): Root directory of the Middleburry 2014 Dataset.
         split (string, optional): The dataset split of scenes, either "train" (default), "test", or "additional"
         use_ambient_views (boolean, optional): Whether to use different expose or lightning views when possible.
             The dataset samples with equal probability between ``[im1.png, im1E.png, im1L.png]``.
@@ -480,7 +480,7 @@ class Middlebury2014Stereo(StereoMatchingDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         calibration: Optional[str] = "perfect",
         use_ambient_views: bool = False,
@@ -576,7 +576,7 @@ def _read_disparity(self, file_path: str) -> Union[Tuple[None, None], Tuple[np.n
         valid_mask = (disparity_map > 0).squeeze(0)  # mask out invalid disparities
         return disparity_map, valid_mask
 
-    def _download_dataset(self, root: str) -> None:
+    def _download_dataset(self, root: Union[str, Path]) -> None:
         base_url = "https://vision.middlebury.edu/stereo/data/scenes2014/zip"
         # train and additional splits have 2 different calibration settings
         root = Path(root) / "Middlebury2014"
@@ -675,7 +675,7 @@ class CREStereo(StereoMatchingDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         transforms: Optional[Callable] = None,
     ) -> None:
         super().__init__(root, transforms)
@@ -757,12 +757,12 @@ class FallingThingsStereo(StereoMatchingDataset):
                     ...
 
     Args:
-        root (string): Root directory where FallingThings is located.
+        root (str or ``pathlib.Path``): Root directory where FallingThings is located.
         variant (string): Which variant to use. Either "single", "mixed", or "both".
         transforms (callable, optional): A function/transform that takes in a sample and returns a transformed version.
     """
 
-    def __init__(self, root: str, variant: str = "single", transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], variant: str = "single", transforms: Optional[Callable] = None) -> None:
         super().__init__(root, transforms)
 
         root = Path(root) / "FallingThings"
@@ -868,7 +868,7 @@ class SceneFlowStereo(StereoMatchingDataset):
                     ...
 
     Args:
-        root (string): Root directory where SceneFlow is located.
+        root (str or ``pathlib.Path``): Root directory where SceneFlow is located.
         variant (string): Which dataset variant to user, "FlyingThings3D" (default), "Monkaa" or "Driving".
         pass_name (string): Which pass to use, "clean" (default), "final" or "both".
         transforms (callable, optional): A function/transform that takes in a sample and returns a transformed version.
@@ -877,7 +877,7 @@ class SceneFlowStereo(StereoMatchingDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         variant: str = "FlyingThings3D",
         pass_name: str = "clean",
         transforms: Optional[Callable] = None,
@@ -973,14 +973,14 @@ class SintelStereo(StereoMatchingDataset):
                         ...
 
     Args:
-        root (string): Root directory where Sintel Stereo is located.
+        root (str or ``pathlib.Path``): Root directory where Sintel Stereo is located.
         pass_name (string): The name of the pass to use, either "final", "clean" or "both".
         transforms (callable, optional): A function/transform that takes in a sample and returns a transformed version.
     """
 
     _has_built_in_disparity_mask = True
 
-    def __init__(self, root: str, pass_name: str = "final", transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], pass_name: str = "final", transforms: Optional[Callable] = None) -> None:
         super().__init__(root, transforms)
 
         verify_str_arg(pass_name, "pass_name", valid_values=("final", "clean", "both"))
@@ -1082,12 +1082,12 @@ class InStereo2k(StereoMatchingDataset):
                     ...
 
     Args:
-        root (string): Root directory where InStereo2k is located.
+        root (str or ``pathlib.Path``): Root directory where InStereo2k is located.
         split (string): Either "train" or "test".
         transforms (callable, optional): A function/transform that takes in a sample and returns a transformed version.
     """
 
-    def __init__(self, root: str, split: str = "train", transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], split: str = "train", transforms: Optional[Callable] = None) -> None:
         super().__init__(root, transforms)
 
         root = Path(root) / "InStereo2k" / split
@@ -1169,14 +1169,14 @@ class ETH3DStereo(StereoMatchingDataset):
                     ...
 
     Args:
-        root (string): Root directory of the ETH3D Dataset.
+        root (str or ``pathlib.Path``): Root directory of the ETH3D Dataset.
         split (string, optional): The dataset split of scenes, either "train" (default) or "test".
         transforms (callable, optional): A function/transform that takes in a sample and returns a transformed version.
     """
 
     _has_built_in_disparity_mask = True
 
-    def __init__(self, root: str, split: str = "train", transforms: Optional[Callable] = None) -> None:
+    def __init__(self, root: Union[str, Path], split: str = "train", transforms: Optional[Callable] = None) -> None:
         super().__init__(root, transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "test"))
diff --git a/torchvision/datasets/caltech.py b/torchvision/datasets/caltech.py
index a1aa7596238..fe4f0fad208 100644
--- a/torchvision/datasets/caltech.py
+++ b/torchvision/datasets/caltech.py
@@ -1,5 +1,6 @@
 import os
 import os.path
+from pathlib import Path
 from typing import Any, Callable, List, Optional, Tuple, Union
 
 from PIL import Image
@@ -16,7 +17,7 @@ class Caltech101(VisionDataset):
         This class needs `scipy <https://docs.scipy.org/doc/>`_ to load target files from `.mat` format.
 
     Args:
-        root (string): Root directory of dataset where directory
+        root (str or ``pathlib.Path``): Root directory of dataset where directory
             ``caltech101`` exists or will be saved to if download is set to True.
         target_type (string or list, optional): Type of target to use, ``category`` or
             ``annotation``. Can also be a list to output a tuple with all specified
@@ -38,7 +39,7 @@ class Caltech101(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         target_type: Union[List[str], str] = "category",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
@@ -153,7 +154,7 @@ class Caltech256(VisionDataset):
     """`Caltech 256 <https://data.caltech.edu/records/20087>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where directory
+        root (str or ``pathlib.Path``): Root directory of dataset where directory
             ``caltech256`` exists or will be saved to if download is set to True.
         transform (callable, optional): A function/transform that takes in a PIL image
             and returns a transformed version. E.g, ``transforms.RandomCrop``
diff --git a/torchvision/datasets/celeba.py b/torchvision/datasets/celeba.py
index 5c4269664eb..147597d3ab3 100644
--- a/torchvision/datasets/celeba.py
+++ b/torchvision/datasets/celeba.py
@@ -1,6 +1,7 @@
 import csv
 import os
 from collections import namedtuple
+from pathlib import Path
 from typing import Any, Callable, List, Optional, Tuple, Union
 
 import PIL
@@ -16,7 +17,7 @@ class CelebA(VisionDataset):
     """`Large-scale CelebFaces Attributes (CelebA) Dataset <http://mmlab.ie.cuhk.edu.hk/projects/CelebA.html>`_ Dataset.
 
     Args:
-        root (string): Root directory where images are downloaded to.
+        root (str or ``pathlib.Path``): Root directory where images are downloaded to.
         split (string): One of {'train', 'valid', 'test', 'all'}.
             Accordingly dataset is selected.
         target_type (string or list, optional): Type of target to use, ``attr``, ``identity``, ``bbox``,
@@ -63,7 +64,7 @@ class CelebA(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         target_type: Union[List[str], str] = "attr",
         transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/cifar.py b/torchvision/datasets/cifar.py
index 4df0306634e..1637670ab91 100644
--- a/torchvision/datasets/cifar.py
+++ b/torchvision/datasets/cifar.py
@@ -1,6 +1,7 @@
 import os.path
 import pickle
-from typing import Any, Callable, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Optional, Tuple, Union
 
 import numpy as np
 from PIL import Image
@@ -13,7 +14,7 @@ class CIFAR10(VisionDataset):
     """`CIFAR10 <https://www.cs.toronto.edu/~kriz/cifar.html>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where directory
+        root (str or ``pathlib.Path``): Root directory of dataset where directory
             ``cifar-10-batches-py`` exists or will be saved to if download is set to True.
         train (bool, optional): If True, creates dataset from training set, otherwise
             creates from test set.
@@ -50,7 +51,7 @@ class CIFAR10(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         train: bool = True,
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/cityscapes.py b/torchvision/datasets/cityscapes.py
index 85544598176..6f7281f2574 100644
--- a/torchvision/datasets/cityscapes.py
+++ b/torchvision/datasets/cityscapes.py
@@ -1,6 +1,7 @@
 import json
 import os
 from collections import namedtuple
+from pathlib import Path
 from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 
 from PIL import Image
@@ -13,7 +14,7 @@ class Cityscapes(VisionDataset):
     """`Cityscapes <http://www.cityscapes-dataset.com/>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where directory ``leftImg8bit``
+        root (str or ``pathlib.Path``): Root directory of dataset where directory ``leftImg8bit``
             and ``gtFine`` or ``gtCoarse`` are located.
         split (string, optional): The image split to use, ``train``, ``test`` or ``val`` if mode="fine"
             otherwise ``train``, ``train_extra`` or ``val``
@@ -103,7 +104,7 @@ class Cityscapes(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         mode: str = "fine",
         target_type: Union[List[str], str] = "instance",
diff --git a/torchvision/datasets/clevr.py b/torchvision/datasets/clevr.py
index f9ee63558fd..328eb7d79da 100644
--- a/torchvision/datasets/clevr.py
+++ b/torchvision/datasets/clevr.py
@@ -1,6 +1,6 @@
 import json
 import pathlib
-from typing import Any, Callable, List, Optional, Tuple
+from typing import Any, Callable, List, Optional, Tuple, Union
 from urllib.parse import urlparse
 
 from PIL import Image
@@ -15,7 +15,7 @@ class CLEVRClassification(VisionDataset):
     The number of objects in a scene are used as label.
 
     Args:
-        root (string): Root directory of dataset where directory ``root/clevr`` exists or will be saved to if download is
+        root (str or ``pathlib.Path``): Root directory of dataset where directory ``root/clevr`` exists or will be saved to if download is
             set to True.
         split (string, optional): The dataset split, supports ``"train"`` (default), ``"val"``, or ``"test"``.
         transform (callable, optional): A function/transform that takes in a PIL image and returns a transformed
@@ -30,7 +30,7 @@ class CLEVRClassification(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, pathlib.Path],
         split: str = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/coco.py b/torchvision/datasets/coco.py
index fd374a8ec24..f3b7be798b2 100644
--- a/torchvision/datasets/coco.py
+++ b/torchvision/datasets/coco.py
@@ -1,5 +1,6 @@
 import os.path
-from typing import Any, Callable, List, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, List, Optional, Tuple, Union
 
 from PIL import Image
 
@@ -12,7 +13,7 @@ class CocoDetection(VisionDataset):
     It requires the `COCO API to be installed <https://github.com/pdollar/coco/tree/master/PythonAPI>`_.
 
     Args:
-        root (string): Root directory where images are downloaded to.
+        root (str or ``pathlib.Path``): Root directory where images are downloaded to.
         annFile (string): Path to json annotation file.
         transform (callable, optional): A function/transform that takes in a PIL image
             and returns a transformed version. E.g, ``transforms.PILToTensor``
@@ -24,7 +25,7 @@ class CocoDetection(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         annFile: str,
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
@@ -67,7 +68,7 @@ class CocoCaptions(CocoDetection):
     It requires the `COCO API to be installed <https://github.com/pdollar/coco/tree/master/PythonAPI>`_.
 
     Args:
-        root (string): Root directory where images are downloaded to.
+        root (str or ``pathlib.Path``): Root directory where images are downloaded to.
         annFile (string): Path to json annotation file.
         transform (callable, optional): A function/transform that  takes in a PIL image
             and returns a transformed version. E.g, ``transforms.PILToTensor``
diff --git a/torchvision/datasets/country211.py b/torchvision/datasets/country211.py
index 906fdfb2065..a0f82ee1226 100644
--- a/torchvision/datasets/country211.py
+++ b/torchvision/datasets/country211.py
@@ -1,5 +1,5 @@
 from pathlib import Path
-from typing import Callable, Optional
+from typing import Callable, Optional, Union
 
 from .folder import ImageFolder
 from .utils import download_and_extract_archive, verify_str_arg
@@ -14,7 +14,7 @@ class Country211(ImageFolder):
     100 test images for each country.
 
     Args:
-        root (string): Root directory of the dataset.
+        root (str or ``pathlib.Path``): Root directory of the dataset.
         split (string, optional): The dataset split, supports ``"train"`` (default), ``"valid"`` and ``"test"``.
         transform (callable, optional): A function/transform that takes in a PIL image and returns a transformed
             version. E.g, ``transforms.RandomCrop``.
@@ -28,7 +28,7 @@ class Country211(ImageFolder):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/dtd.py b/torchvision/datasets/dtd.py
index b4e045e7c20..71c556bd201 100644
--- a/torchvision/datasets/dtd.py
+++ b/torchvision/datasets/dtd.py
@@ -1,6 +1,6 @@
 import os
 import pathlib
-from typing import Any, Callable, Optional, Tuple
+from typing import Any, Callable, Optional, Tuple, Union
 
 import PIL.Image
 
@@ -12,7 +12,7 @@ class DTD(VisionDataset):
     """`Describable Textures Dataset (DTD) <https://www.robots.ox.ac.uk/~vgg/data/dtd/>`_.
 
     Args:
-        root (string): Root directory of the dataset.
+        root (str or ``pathlib.Path``): Root directory of the dataset.
         split (string, optional): The dataset split, supports ``"train"`` (default), ``"val"``, or ``"test"``.
         partition (int, optional): The dataset partition. Should be ``1 <= partition <= 10``. Defaults to ``1``.
 
@@ -34,7 +34,7 @@ class DTD(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, pathlib.Path],
         split: str = "train",
         partition: int = 1,
         transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/eurosat.py b/torchvision/datasets/eurosat.py
index 2ae8c073202..3f490b11902 100644
--- a/torchvision/datasets/eurosat.py
+++ b/torchvision/datasets/eurosat.py
@@ -1,5 +1,6 @@
 import os
-from typing import Callable, Optional
+from pathlib import Path
+from typing import Callable, Optional, Union
 
 from .folder import ImageFolder
 from .utils import download_and_extract_archive
@@ -9,7 +10,7 @@ class EuroSAT(ImageFolder):
     """RGB version of the `EuroSAT <https://github.com/phelber/eurosat>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where ``root/eurosat`` exists.
+        root (str or ``pathlib.Path``): Root directory of dataset where ``root/eurosat`` exists.
         transform (callable, optional): A function/transform that takes in a PIL image
             and returns a transformed version. E.g, ``transforms.RandomCrop``
         target_transform (callable, optional): A function/transform that takes in the
@@ -21,7 +22,7 @@ class EuroSAT(ImageFolder):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
         download: bool = False,
diff --git a/torchvision/datasets/fer2013.py b/torchvision/datasets/fer2013.py
index 90bf5f304cd..057fe695a13 100644
--- a/torchvision/datasets/fer2013.py
+++ b/torchvision/datasets/fer2013.py
@@ -1,6 +1,6 @@
 import csv
 import pathlib
-from typing import Any, Callable, Optional, Tuple
+from typing import Any, Callable, Optional, Tuple, Union
 
 import torch
 from PIL import Image
@@ -14,7 +14,7 @@ class FER2013(VisionDataset):
     <https://www.kaggle.com/c/challenges-in-representation-learning-facial-expression-recognition-challenge>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where directory
+        root (str or ``pathlib.Path``): Root directory of dataset where directory
             ``root/fer2013`` exists.
         split (string, optional): The dataset split, supports ``"train"`` (default), or ``"test"``.
         transform (callable, optional): A function/transform that takes in a PIL image and returns a transformed
@@ -29,7 +29,7 @@ class FER2013(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, pathlib.Path],
         split: str = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/fgvc_aircraft.py b/torchvision/datasets/fgvc_aircraft.py
index e4b071712c3..bbf4e970a78 100644
--- a/torchvision/datasets/fgvc_aircraft.py
+++ b/torchvision/datasets/fgvc_aircraft.py
@@ -1,7 +1,8 @@
 from __future__ import annotations
 
 import os
-from typing import Any, Callable, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Optional, Tuple, Union
 
 import PIL.Image
 
@@ -23,7 +24,7 @@ class FGVCAircraft(VisionDataset):
     - ``manufacturer``, e.g. Boeing. The dataset comprises 30 different manufacturers.
 
     Args:
-        root (string): Root directory of the FGVC Aircraft dataset.
+        root (str or ``pathlib.Path``): Root directory of the FGVC Aircraft dataset.
         split (string, optional): The dataset split, supports ``train``, ``val``,
             ``trainval`` and ``test``.
         annotation_level (str, optional): The annotation level, supports ``variant``,
@@ -41,7 +42,7 @@ class FGVCAircraft(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "trainval",
         annotation_level: str = "variant",
         transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/flickr.py b/torchvision/datasets/flickr.py
index 0047a12268b..1021309db05 100644
--- a/torchvision/datasets/flickr.py
+++ b/torchvision/datasets/flickr.py
@@ -2,7 +2,8 @@
 import os
 from collections import defaultdict
 from html.parser import HTMLParser
-from typing import Any, Callable, Dict, List, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 
 from PIL import Image
 
@@ -12,7 +13,7 @@
 class Flickr8kParser(HTMLParser):
     """Parser for extracting captions from the Flickr8k dataset web page."""
 
-    def __init__(self, root: str) -> None:
+    def __init__(self, root: Union[str, Path]) -> None:
         super().__init__()
 
         self.root = root
@@ -56,7 +57,7 @@ class Flickr8k(VisionDataset):
     """`Flickr8k Entities <http://hockenmaier.cs.illinois.edu/8k-pictures.html>`_ Dataset.
 
     Args:
-        root (string): Root directory where images are downloaded to.
+        root (str or ``pathlib.Path``): Root directory where images are downloaded to.
         ann_file (string): Path to annotation file.
         transform (callable, optional): A function/transform that takes in a PIL image
             and returns a transformed version. E.g, ``transforms.PILToTensor``
@@ -66,7 +67,7 @@ class Flickr8k(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         ann_file: str,
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
@@ -112,7 +113,7 @@ class Flickr30k(VisionDataset):
     """`Flickr30k Entities <https://bryanplummer.com/Flickr30kEntities/>`_ Dataset.
 
     Args:
-        root (string): Root directory where images are downloaded to.
+        root (str or ``pathlib.Path``): Root directory where images are downloaded to.
         ann_file (string): Path to annotation file.
         transform (callable, optional): A function/transform that takes in a PIL image
             and returns a transformed version. E.g, ``transforms.PILToTensor``
diff --git a/torchvision/datasets/flowers102.py b/torchvision/datasets/flowers102.py
index 41966f336d2..07f403702f5 100644
--- a/torchvision/datasets/flowers102.py
+++ b/torchvision/datasets/flowers102.py
@@ -1,5 +1,5 @@
 from pathlib import Path
-from typing import Any, Callable, Optional, Tuple
+from typing import Any, Callable, Optional, Tuple, Union
 
 import PIL.Image
 
@@ -22,7 +22,7 @@ class Flowers102(VisionDataset):
     have large variations within the category, and several very similar categories.
 
     Args:
-        root (string): Root directory of the dataset.
+        root (str or ``pathlib.Path``): Root directory of the dataset.
         split (string, optional): The dataset split, supports ``"train"`` (default), ``"val"``, or ``"test"``.
         transform (callable, optional): A function/transform that takes in a PIL image and returns a
             transformed version. E.g, ``transforms.RandomCrop``.
@@ -42,7 +42,7 @@ class Flowers102(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/folder.py b/torchvision/datasets/folder.py
index 3a38d556287..9ee06b6a650 100644
--- a/torchvision/datasets/folder.py
+++ b/torchvision/datasets/folder.py
@@ -1,5 +1,6 @@
 import os
 import os.path
+from pathlib import Path
 from typing import Any, Callable, cast, Dict, List, Optional, Tuple, Union
 
 from PIL import Image
@@ -32,7 +33,7 @@ def is_image_file(filename: str) -> bool:
     return has_file_allowed_extension(filename, IMG_EXTENSIONS)
 
 
-def find_classes(directory: str) -> Tuple[List[str], Dict[str, int]]:
+def find_classes(directory: Union[str, Path]) -> Tuple[List[str], Dict[str, int]]:
     """Finds the class folders in a dataset.
 
     See :class:`DatasetFolder` for details.
@@ -46,7 +47,7 @@ def find_classes(directory: str) -> Tuple[List[str], Dict[str, int]]:
 
 
 def make_dataset(
-    directory: str,
+    directory: Union[str, Path],
     class_to_idx: Optional[Dict[str, int]] = None,
     extensions: Optional[Union[str, Tuple[str, ...]]] = None,
     is_valid_file: Optional[Callable[[str], bool]] = None,
@@ -112,7 +113,7 @@ class DatasetFolder(VisionDataset):
     :meth:`find_classes` method.
 
     Args:
-        root (string): Root directory path.
+        root (str or ``pathlib.Path``): Root directory path.
         loader (callable): A function to load a sample given its path.
         extensions (tuple[string]): A list of allowed extensions.
             both extensions and is_valid_file should not be passed.
@@ -136,7 +137,7 @@ class DatasetFolder(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         loader: Callable[[str], Any],
         extensions: Optional[Tuple[str, ...]] = None,
         transform: Optional[Callable] = None,
@@ -164,7 +165,7 @@ def __init__(
 
     @staticmethod
     def make_dataset(
-        directory: str,
+        directory: Union[str, Path],
         class_to_idx: Dict[str, int],
         extensions: Optional[Tuple[str, ...]] = None,
         is_valid_file: Optional[Callable[[str], bool]] = None,
@@ -203,7 +204,7 @@ def make_dataset(
             directory, class_to_idx, extensions=extensions, is_valid_file=is_valid_file, allow_empty=allow_empty
         )
 
-    def find_classes(self, directory: str) -> Tuple[List[str], Dict[str, int]]:
+    def find_classes(self, directory: Union[str, Path]) -> Tuple[List[str], Dict[str, int]]:
         """Find the class folders in a dataset structured as follows::
 
             directory/
@@ -298,7 +299,7 @@ class ImageFolder(DatasetFolder):
     the same methods can be overridden to customize the dataset.
 
     Args:
-        root (string): Root directory path.
+        root (str or ``pathlib.Path``): Root directory path.
         transform (callable, optional): A function/transform that takes in a PIL image
             and returns a transformed version. E.g, ``transforms.RandomCrop``
         target_transform (callable, optional): A function/transform that takes in the
diff --git a/torchvision/datasets/food101.py b/torchvision/datasets/food101.py
index fa38abd7238..f734787c1bf 100644
--- a/torchvision/datasets/food101.py
+++ b/torchvision/datasets/food101.py
@@ -1,6 +1,6 @@
 import json
 from pathlib import Path
-from typing import Any, Callable, Optional, Tuple
+from typing import Any, Callable, Optional, Tuple, Union
 
 import PIL.Image
 
@@ -19,7 +19,7 @@ class Food101(VisionDataset):
 
 
     Args:
-        root (string): Root directory of the dataset.
+        root (str or ``pathlib.Path``): Root directory of the dataset.
         split (string, optional): The dataset split, supports ``"train"`` (default) and ``"test"``.
         transform (callable, optional): A function/transform that takes in a PIL image and returns a transformed
             version. E.g, ``transforms.RandomCrop``.
@@ -34,7 +34,7 @@ class Food101(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/gtsrb.py b/torchvision/datasets/gtsrb.py
index ec746fbf686..a3d012c70b2 100644
--- a/torchvision/datasets/gtsrb.py
+++ b/torchvision/datasets/gtsrb.py
@@ -1,6 +1,6 @@
 import csv
 import pathlib
-from typing import Any, Callable, Optional, Tuple
+from typing import Any, Callable, Optional, Tuple, Union
 
 import PIL
 
@@ -13,7 +13,7 @@ class GTSRB(VisionDataset):
     """`German Traffic Sign Recognition Benchmark (GTSRB) <https://benchmark.ini.rub.de/>`_ Dataset.
 
     Args:
-        root (string): Root directory of the dataset.
+        root (str or ``pathlib.Path``): Root directory of the dataset.
         split (string, optional): The dataset split, supports ``"train"`` (default), or ``"test"``.
         transform (callable, optional): A function/transform that takes in a PIL image and returns a transformed
             version. E.g, ``transforms.RandomCrop``.
@@ -25,7 +25,7 @@ class GTSRB(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, pathlib.Path],
         split: str = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/hmdb51.py b/torchvision/datasets/hmdb51.py
index a58ddc293d9..8377e40d57c 100644
--- a/torchvision/datasets/hmdb51.py
+++ b/torchvision/datasets/hmdb51.py
@@ -1,6 +1,7 @@
 import glob
 import os
-from typing import Any, Callable, Dict, List, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 
 from torch import Tensor
 
@@ -28,7 +29,7 @@ class HMDB51(VisionDataset):
     Internally, it uses a VideoClips object to handle clip creation.
 
     Args:
-        root (string): Root directory of the HMDB51 Dataset.
+        root (str or ``pathlib.Path``): Root directory of the HMDB51 Dataset.
         annotation_path (str): Path to the folder containing the split files.
         frames_per_clip (int): Number of frames in a clip.
         step_between_clips (int): Number of frames between each clip.
@@ -59,7 +60,7 @@ class HMDB51(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         annotation_path: str,
         frames_per_clip: int,
         step_between_clips: int = 1,
diff --git a/torchvision/datasets/imagenet.py b/torchvision/datasets/imagenet.py
index 290ae8f157f..d7caf328d2b 100644
--- a/torchvision/datasets/imagenet.py
+++ b/torchvision/datasets/imagenet.py
@@ -2,7 +2,8 @@
 import shutil
 import tempfile
 from contextlib import contextmanager
-from typing import Any, Dict, Iterator, List, Optional, Tuple
+from pathlib import Path
+from typing import Any, Dict, Iterator, List, Optional, Tuple, Union
 
 import torch
 
@@ -28,7 +29,7 @@ class ImageNet(ImageFolder):
         or ``ILSVRC2012_img_val.tar`` based on ``split`` in the root directory.
 
     Args:
-        root (string): Root directory of the ImageNet Dataset.
+        root (str or ``pathlib.Path``): Root directory of the ImageNet Dataset.
         split (string, optional): The dataset split, supports ``train``, or ``val``.
         transform (callable, optional): A function/transform that takes in a PIL image
             and returns a transformed version. E.g, ``transforms.RandomCrop``
@@ -45,7 +46,7 @@ class ImageNet(ImageFolder):
         targets (list): The class_index value for each image in the dataset
     """
 
-    def __init__(self, root: str, split: str = "train", **kwargs: Any) -> None:
+    def __init__(self, root: Union[str, Path], split: str = "train", **kwargs: Any) -> None:
         root = self.root = os.path.expanduser(root)
         self.split = verify_str_arg(split, "split", ("train", "val"))
 
@@ -78,7 +79,7 @@ def extra_repr(self) -> str:
         return "Split: {split}".format(**self.__dict__)
 
 
-def load_meta_file(root: str, file: Optional[str] = None) -> Tuple[Dict[str, str], List[str]]:
+def load_meta_file(root: Union[str, Path], file: Optional[str] = None) -> Tuple[Dict[str, str], List[str]]:
     if file is None:
         file = META_FILE
     file = os.path.join(root, file)
@@ -93,7 +94,7 @@ def load_meta_file(root: str, file: Optional[str] = None) -> Tuple[Dict[str, str
         raise RuntimeError(msg.format(file, root))
 
 
-def _verify_archive(root: str, file: str, md5: str) -> None:
+def _verify_archive(root: Union[str, Path], file: str, md5: str) -> None:
     if not check_integrity(os.path.join(root, file), md5):
         msg = (
             "The archive {} is not present in the root directory or is corrupted. "
@@ -102,12 +103,12 @@ def _verify_archive(root: str, file: str, md5: str) -> None:
         raise RuntimeError(msg.format(file, root))
 
 
-def parse_devkit_archive(root: str, file: Optional[str] = None) -> None:
+def parse_devkit_archive(root: Union[str, Path], file: Optional[str] = None) -> None:
     """Parse the devkit archive of the ImageNet2012 classification dataset and save
     the meta information in a binary file.
 
     Args:
-        root (str): Root directory containing the devkit archive
+        root (str or ``pathlib.Path``): Root directory containing the devkit archive
         file (str, optional): Name of devkit archive. Defaults to
             'ILSVRC2012_devkit_t12.tar.gz'
     """
@@ -156,12 +157,12 @@ def get_tmp_dir() -> Iterator[str]:
         torch.save((wnid_to_classes, val_wnids), os.path.join(root, META_FILE))
 
 
-def parse_train_archive(root: str, file: Optional[str] = None, folder: str = "train") -> None:
+def parse_train_archive(root: Union[str, Path], file: Optional[str] = None, folder: str = "train") -> None:
     """Parse the train images archive of the ImageNet2012 classification dataset and
     prepare it for usage with the ImageNet dataset.
 
     Args:
-        root (str): Root directory containing the train images archive
+        root (str or ``pathlib.Path``): Root directory containing the train images archive
         file (str, optional): Name of train images archive. Defaults to
             'ILSVRC2012_img_train.tar'
         folder (str, optional): Optional name for train images folder. Defaults to
@@ -183,13 +184,13 @@ def parse_train_archive(root: str, file: Optional[str] = None, folder: str = "tr
 
 
 def parse_val_archive(
-    root: str, file: Optional[str] = None, wnids: Optional[List[str]] = None, folder: str = "val"
+    root: Union[str, Path], file: Optional[str] = None, wnids: Optional[List[str]] = None, folder: str = "val"
 ) -> None:
     """Parse the validation images archive of the ImageNet2012 classification dataset
     and prepare it for usage with the ImageNet dataset.
 
     Args:
-        root (str): Root directory containing the validation images archive
+        root (str or ``pathlib.Path``): Root directory containing the validation images archive
         file (str, optional): Name of validation images archive. Defaults to
             'ILSVRC2012_img_val.tar'
         wnids (list, optional): List of WordNet IDs of the validation images. If None
diff --git a/torchvision/datasets/imagenette.py b/torchvision/datasets/imagenette.py
index bf0835ce0ab..05da537891b 100644
--- a/torchvision/datasets/imagenette.py
+++ b/torchvision/datasets/imagenette.py
@@ -1,5 +1,5 @@
 from pathlib import Path
-from typing import Any, Callable, Optional, Tuple
+from typing import Any, Callable, Optional, Tuple, Union
 
 from PIL import Image
 
@@ -12,7 +12,7 @@ class Imagenette(VisionDataset):
     """`Imagenette <https://github.com/fastai/imagenette#imagenette-1>`_ image classification dataset.
 
     Args:
-        root (string): Root directory of the Imagenette dataset.
+        root (str or ``pathlib.Path``): Root directory of the Imagenette dataset.
         split (string, optional): The dataset split. Supports ``"train"`` (default), and ``"val"``.
         size (string, optional): The image size. Supports ``"full"`` (default), ``"320px"``, and ``"160px"``.
         download (bool, optional): If ``True``, downloads the dataset components and places them in ``root``. Already
@@ -48,7 +48,7 @@ class Imagenette(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         size: str = "full",
         download=False,
diff --git a/torchvision/datasets/inaturalist.py b/torchvision/datasets/inaturalist.py
index 84d6907b485..68f9a77f56a 100644
--- a/torchvision/datasets/inaturalist.py
+++ b/torchvision/datasets/inaturalist.py
@@ -1,5 +1,6 @@
 import os
 import os.path
+from pathlib import Path
 from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 
 from PIL import Image
@@ -32,7 +33,7 @@ class INaturalist(VisionDataset):
     """`iNaturalist <https://github.com/visipedia/inat_comp>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where the image files are stored.
+        root (str or ``pathlib.Path``): Root directory of dataset where the image files are stored.
             This class does not require/use annotation files.
         version (string, optional): Which version of the dataset to download/use. One of
             '2017', '2018', '2019', '2021_train', '2021_train_mini', '2021_valid'.
@@ -65,7 +66,7 @@ class INaturalist(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         version: str = "2021_train",
         target_type: Union[List[str], str] = "full",
         transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/kinetics.py b/torchvision/datasets/kinetics.py
index a5fda53aa90..42d32533953 100644
--- a/torchvision/datasets/kinetics.py
+++ b/torchvision/datasets/kinetics.py
@@ -5,7 +5,8 @@
 from functools import partial
 from multiprocessing import Pool
 from os import path
-from typing import Any, Callable, Dict, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Dict, Optional, Tuple, Union
 
 from torch import Tensor
 
@@ -35,7 +36,7 @@ class Kinetics(VisionDataset):
     frames in a video might be present.
 
     Args:
-        root (string): Root directory of the Kinetics Dataset.
+        root (str or ``pathlib.Path``): Root directory of the Kinetics Dataset.
             Directory should be structured as follows:
             .. code::
 
@@ -90,7 +91,7 @@ class Kinetics(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         frames_per_clip: int,
         num_classes: str = "400",
         split: str = "train",
diff --git a/torchvision/datasets/kitti.py b/torchvision/datasets/kitti.py
index 37b99006c75..69e603c76f2 100644
--- a/torchvision/datasets/kitti.py
+++ b/torchvision/datasets/kitti.py
@@ -1,6 +1,7 @@
 import csv
 import os
-from typing import Any, Callable, List, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, List, Optional, Tuple, Union
 
 from PIL import Image
 
@@ -14,7 +15,7 @@ class Kitti(VisionDataset):
     It corresponds to the "left color images of object" dataset, for object detection.
 
     Args:
-        root (string): Root directory where images are downloaded to.
+        root (str or ``pathlib.Path``): Root directory where images are downloaded to.
             Expects the following folder structure if download=False:
 
             .. code::
@@ -51,7 +52,7 @@ class Kitti(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         train: bool = True,
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/lfw.py b/torchvision/datasets/lfw.py
index f46c4e4c872..69f1edaf72f 100644
--- a/torchvision/datasets/lfw.py
+++ b/torchvision/datasets/lfw.py
@@ -1,4 +1,5 @@
 import os
+from pathlib import Path
 from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 
 from PIL import Image
@@ -31,7 +32,7 @@ class _LFW(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str,
         image_set: str,
         view: str,
@@ -95,7 +96,7 @@ class LFWPeople(_LFW):
     """`LFW <http://vis-www.cs.umass.edu/lfw/>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where directory
+        root (str or ``pathlib.Path``): Root directory of dataset where directory
             ``lfw-py`` exists or will be saved to if download is set to True.
         split (string, optional): The image split to use. Can be one of ``train``, ``test``,
             ``10fold`` (default).
@@ -177,7 +178,7 @@ class LFWPairs(_LFW):
     """`LFW <http://vis-www.cs.umass.edu/lfw/>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where directory
+        root (str or ``pathlib.Path``): Root directory of dataset where directory
             ``lfw-py`` exists or will be saved to if download is set to True.
         split (string, optional): The image split to use. Can be one of ``train``, ``test``,
             ``10fold``. Defaults to ``10fold``.
diff --git a/torchvision/datasets/lsun.py b/torchvision/datasets/lsun.py
index 6fab66ff11e..a2f5e18b991 100644
--- a/torchvision/datasets/lsun.py
+++ b/torchvision/datasets/lsun.py
@@ -3,6 +3,7 @@
 import pickle
 import string
 from collections.abc import Iterable
+from pathlib import Path
 from typing import Any, Callable, cast, List, Optional, Tuple, Union
 
 from PIL import Image
@@ -60,7 +61,7 @@ class LSUN(VisionDataset):
     ``pip install lmdb``
 
     Args:
-        root (string): Root directory for the database files.
+        root (str or ``pathlib.Path``): Root directory for the database files.
         classes (string or list): One of {'train', 'val', 'test'} or a list of
             categories to load. e,g. ['bedroom_train', 'church_outdoor_train'].
         transform (callable, optional): A function/transform that takes in a PIL image
@@ -71,7 +72,7 @@ class LSUN(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         classes: Union[str, List[str]] = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/mnist.py b/torchvision/datasets/mnist.py
index fce8583985c..a2389d598e6 100644
--- a/torchvision/datasets/mnist.py
+++ b/torchvision/datasets/mnist.py
@@ -5,7 +5,8 @@
 import string
 import sys
 import warnings
-from typing import Any, Callable, Dict, List, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 from urllib.error import URLError
 
 import numpy as np
@@ -20,7 +21,7 @@ class MNIST(VisionDataset):
     """`MNIST <http://yann.lecun.com/exdb/mnist/>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where ``MNIST/raw/train-images-idx3-ubyte``
+        root (str or ``pathlib.Path``): Root directory of dataset where ``MNIST/raw/train-images-idx3-ubyte``
             and  ``MNIST/raw/t10k-images-idx3-ubyte`` exist.
         train (bool, optional): If True, creates dataset from ``train-images-idx3-ubyte``,
             otherwise from ``t10k-images-idx3-ubyte``.
@@ -82,7 +83,7 @@ def test_data(self):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         train: bool = True,
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
@@ -203,7 +204,7 @@ class FashionMNIST(MNIST):
     """`Fashion-MNIST <https://github.com/zalandoresearch/fashion-mnist>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where ``FashionMNIST/raw/train-images-idx3-ubyte``
+        root (str or ``pathlib.Path``): Root directory of dataset where ``FashionMNIST/raw/train-images-idx3-ubyte``
             and  ``FashionMNIST/raw/t10k-images-idx3-ubyte`` exist.
         train (bool, optional): If True, creates dataset from ``train-images-idx3-ubyte``,
             otherwise from ``t10k-images-idx3-ubyte``.
@@ -231,7 +232,7 @@ class KMNIST(MNIST):
     """`Kuzushiji-MNIST <https://github.com/rois-codh/kmnist>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where ``KMNIST/raw/train-images-idx3-ubyte``
+        root (str or ``pathlib.Path``): Root directory of dataset where ``KMNIST/raw/train-images-idx3-ubyte``
             and  ``KMNIST/raw/t10k-images-idx3-ubyte`` exist.
         train (bool, optional): If True, creates dataset from ``train-images-idx3-ubyte``,
             otherwise from ``t10k-images-idx3-ubyte``.
@@ -259,7 +260,7 @@ class EMNIST(MNIST):
     """`EMNIST <https://www.westernsydney.edu.au/bens/home/reproducible_research/emnist>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where ``EMNIST/raw/train-images-idx3-ubyte``
+        root (str or ``pathlib.Path``): Root directory of dataset where ``EMNIST/raw/train-images-idx3-ubyte``
             and  ``EMNIST/raw/t10k-images-idx3-ubyte`` exist.
         split (string): The dataset has 6 different splits: ``byclass``, ``bymerge``,
             ``balanced``, ``letters``, ``digits`` and ``mnist``. This argument specifies
@@ -290,7 +291,7 @@ class EMNIST(MNIST):
         "mnist": list(string.digits),
     }
 
-    def __init__(self, root: str, split: str, **kwargs: Any) -> None:
+    def __init__(self, root: Union[str, Path], split: str, **kwargs: Any) -> None:
         self.split = verify_str_arg(split, "split", self.splits)
         self.training_file = self._training_file(split)
         self.test_file = self._test_file(split)
@@ -343,7 +344,7 @@ class QMNIST(MNIST):
     """`QMNIST <https://github.com/facebookresearch/qmnist>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset whose ``raw``
+        root (str or ``pathlib.Path``): Root directory of dataset whose ``raw``
             subdir contains binary files of the datasets.
         what (string,optional): Can be 'train', 'test', 'test10k',
             'test50k', or 'nist' for respectively the mnist compatible
@@ -416,7 +417,7 @@ class QMNIST(MNIST):
     ]
 
     def __init__(
-        self, root: str, what: Optional[str] = None, compat: bool = True, train: bool = True, **kwargs: Any
+        self, root: Union[str, Path], what: Optional[str] = None, compat: bool = True, train: bool = True, **kwargs: Any
     ) -> None:
         if what is None:
             what = "train" if train else "test"
diff --git a/torchvision/datasets/moving_mnist.py b/torchvision/datasets/moving_mnist.py
index 9dea36e4223..d02811762b8 100644
--- a/torchvision/datasets/moving_mnist.py
+++ b/torchvision/datasets/moving_mnist.py
@@ -1,5 +1,6 @@
 import os.path
-from typing import Callable, Optional
+from pathlib import Path
+from typing import Callable, Optional, Union
 
 import numpy as np
 import torch
@@ -11,7 +12,7 @@ class MovingMNIST(VisionDataset):
     """`MovingMNIST <http://www.cs.toronto.edu/~nitish/unsupervised_video/>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where ``MovingMNIST/mnist_test_seq.npy`` exists.
+        root (str or ``pathlib.Path``): Root directory of dataset where ``MovingMNIST/mnist_test_seq.npy`` exists.
         split (string, optional): The dataset split, supports ``None`` (default), ``"train"`` and ``"test"``.
             If ``split=None``, the full data is returned.
         split_ratio (int, optional): The split ratio of number of frames. If ``split="train"``, the first split
@@ -28,7 +29,7 @@ class MovingMNIST(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: Optional[str] = None,
         split_ratio: int = 10,
         download: bool = False,
diff --git a/torchvision/datasets/omniglot.py b/torchvision/datasets/omniglot.py
index 42717529a47..c02cf91234a 100644
--- a/torchvision/datasets/omniglot.py
+++ b/torchvision/datasets/omniglot.py
@@ -1,5 +1,6 @@
 from os.path import join
-from typing import Any, Callable, List, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, List, Optional, Tuple, Union
 
 from PIL import Image
 
@@ -11,7 +12,7 @@ class Omniglot(VisionDataset):
     """`Omniglot <https://github.com/brendenlake/omniglot>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where directory
+        root (str or ``pathlib.Path``): Root directory of dataset where directory
             ``omniglot-py`` exists.
         background (bool, optional): If True, creates dataset from the "background" set, otherwise
             creates from the "evaluation" set. This terminology is defined by the authors.
@@ -33,7 +34,7 @@ class Omniglot(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         background: bool = True,
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/oxford_iiit_pet.py b/torchvision/datasets/oxford_iiit_pet.py
index e4022251a8d..9fe78901626 100644
--- a/torchvision/datasets/oxford_iiit_pet.py
+++ b/torchvision/datasets/oxford_iiit_pet.py
@@ -13,7 +13,7 @@ class OxfordIIITPet(VisionDataset):
     """`Oxford-IIIT Pet Dataset   <https://www.robots.ox.ac.uk/~vgg/data/pets/>`_.
 
     Args:
-        root (string): Root directory of the dataset.
+        root (str or ``pathlib.Path``): Root directory of the dataset.
         split (string, optional): The dataset split, supports ``"trainval"`` (default) or ``"test"``.
         target_types (string, sequence of strings, optional): Types of target to use. Can be ``category`` (default) or
             ``segmentation``. Can also be a list to output a tuple with all specified target types. The types represent:
@@ -38,7 +38,7 @@ class OxfordIIITPet(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, pathlib.Path],
         split: str = "trainval",
         target_types: Union[Sequence[str], str] = "category",
         transforms: Optional[Callable] = None,
diff --git a/torchvision/datasets/pcam.py b/torchvision/datasets/pcam.py
index 0268e0c9a10..8849e0ea39d 100644
--- a/torchvision/datasets/pcam.py
+++ b/torchvision/datasets/pcam.py
@@ -1,5 +1,5 @@
 import pathlib
-from typing import Any, Callable, Optional, Tuple
+from typing import Any, Callable, Optional, Tuple, Union
 
 from PIL import Image
 
@@ -18,7 +18,7 @@ class PCAM(VisionDataset):
     This dataset requires the ``h5py`` package which you can install with ``pip install h5py``.
 
     Args:
-         root (string): Root directory of the dataset.
+         root (str or ``pathlib.Path``): Root directory of the dataset.
          split (string, optional): The dataset split, supports ``"train"`` (default), ``"test"`` or ``"val"``.
          transform (callable, optional): A function/transform that takes in a PIL image and returns a transformed
              version. E.g, ``transforms.RandomCrop``.
@@ -72,7 +72,7 @@ class PCAM(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, pathlib.Path],
         split: str = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/phototour.py b/torchvision/datasets/phototour.py
index 5c7dd6b66da..fd2466a3d36 100644
--- a/torchvision/datasets/phototour.py
+++ b/torchvision/datasets/phototour.py
@@ -1,4 +1,5 @@
 import os
+from pathlib import Path
 from typing import Any, Callable, List, Optional, Tuple, Union
 
 import numpy as np
@@ -24,7 +25,7 @@ class PhotoTour(VisionDataset):
 
 
     Args:
-        root (string): Root directory where images are.
+        root (str or ``pathlib.Path``): Root directory where images are.
         name (string): Name of the dataset to load.
         transform (callable, optional): A function/transform that takes in a PIL image
             and returns a transformed version.
@@ -87,7 +88,12 @@ class PhotoTour(VisionDataset):
     matches_files = "m50_100000_100000_0.txt"
 
     def __init__(
-        self, root: str, name: str, train: bool = True, transform: Optional[Callable] = None, download: bool = False
+        self,
+        root: Union[str, Path],
+        name: str,
+        train: bool = True,
+        transform: Optional[Callable] = None,
+        download: bool = False,
     ) -> None:
         super().__init__(root, transform=transform)
         self.name = name
diff --git a/torchvision/datasets/places365.py b/torchvision/datasets/places365.py
index b98fc5e33b4..98966e1dc2f 100644
--- a/torchvision/datasets/places365.py
+++ b/torchvision/datasets/places365.py
@@ -1,6 +1,7 @@
 import os
 from os import path
-from typing import Any, Callable, Dict, List, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 from urllib.parse import urljoin
 
 from .folder import default_loader
@@ -12,7 +13,7 @@ class Places365(VisionDataset):
     r"""`Places365 <http://places2.csail.mit.edu/index.html>`_ classification dataset.
 
     Args:
-        root (string): Root directory of the Places365 dataset.
+        root (str or ``pathlib.Path``): Root directory of the Places365 dataset.
         split (string, optional): The dataset split. Can be one of ``train-standard`` (default), ``train-challenge``,
             ``val``.
         small (bool, optional): If ``True``, uses the small images, i.e. resized to 256 x 256 pixels, instead of the
@@ -62,7 +63,7 @@ class Places365(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train-standard",
         small: bool = False,
         download: bool = False,
diff --git a/torchvision/datasets/rendered_sst2.py b/torchvision/datasets/rendered_sst2.py
index 6fa09b677bb..48b0ddfc4fb 100644
--- a/torchvision/datasets/rendered_sst2.py
+++ b/torchvision/datasets/rendered_sst2.py
@@ -1,5 +1,5 @@
 from pathlib import Path
-from typing import Any, Callable, Optional, Tuple
+from typing import Any, Callable, Optional, Tuple, Union
 
 import PIL.Image
 
@@ -20,7 +20,7 @@ class RenderedSST2(VisionDataset):
     (444 positive and 428 negative), and a test split containing 1821 images (909 positive and 912 negative).
 
     Args:
-        root (string): Root directory of the dataset.
+        root (str or ``pathlib.Path``): Root directory of the dataset.
         split (string, optional): The dataset split, supports ``"train"`` (default), `"val"` and ``"test"``.
         transform (callable, optional): A function/transform that takes in a PIL image and returns a transformed
             version. E.g, ``transforms.RandomCrop``.
@@ -35,7 +35,7 @@ class RenderedSST2(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/sbd.py b/torchvision/datasets/sbd.py
index 8399d025b1b..7f245675b2d 100644
--- a/torchvision/datasets/sbd.py
+++ b/torchvision/datasets/sbd.py
@@ -1,6 +1,7 @@
 import os
 import shutil
-from typing import Any, Callable, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Optional, Tuple, Union
 
 import numpy as np
 from PIL import Image
@@ -27,7 +28,7 @@ class SBDataset(VisionDataset):
         This class needs `scipy <https://docs.scipy.org/doc/>`_ to load target files from `.mat` format.
 
     Args:
-        root (string): Root directory of the Semantic Boundaries Dataset
+        root (str or ``pathlib.Path``): Root directory of the Semantic Boundaries Dataset
         image_set (string, optional): Select the image_set to use, ``train``, ``val`` or ``train_noval``.
             Image set ``train_noval`` excludes VOC 2012 val images.
         mode (string, optional): Select target type. Possible values 'boundaries' or 'segmentation'.
@@ -51,7 +52,7 @@ class SBDataset(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         image_set: str = "train",
         mode: str = "boundaries",
         download: bool = False,
diff --git a/torchvision/datasets/sbu.py b/torchvision/datasets/sbu.py
index ee90eeb64ae..3c349370a12 100644
--- a/torchvision/datasets/sbu.py
+++ b/torchvision/datasets/sbu.py
@@ -1,5 +1,6 @@
 import os
-from typing import Any, Callable, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Optional, Tuple, Union
 
 from PIL import Image
 
@@ -11,7 +12,7 @@ class SBU(VisionDataset):
     """`SBU Captioned Photo <http://www.cs.virginia.edu/~vicente/sbucaptions/>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where tarball
+        root (str or ``pathlib.Path``): Root directory of dataset where tarball
             ``SBUCaptionedPhotoDataset.tar.gz`` exists.
         transform (callable, optional): A function/transform that takes in a PIL image
             and returns a transformed version. E.g, ``transforms.RandomCrop``
@@ -28,7 +29,7 @@ class SBU(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
         download: bool = True,
diff --git a/torchvision/datasets/semeion.py b/torchvision/datasets/semeion.py
index 0703d267ad5..d0344c74241 100644
--- a/torchvision/datasets/semeion.py
+++ b/torchvision/datasets/semeion.py
@@ -1,5 +1,6 @@
 import os.path
-from typing import Any, Callable, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Optional, Tuple, Union
 
 import numpy as np
 from PIL import Image
@@ -12,7 +13,7 @@ class SEMEION(VisionDataset):
     r"""`SEMEION <http://archive.ics.uci.edu/ml/datasets/semeion+handwritten+digit>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where directory
+        root (str or ``pathlib.Path``): Root directory of dataset where directory
             ``semeion.py`` exists.
         transform (callable, optional): A function/transform that takes in a PIL image
             and returns a transformed version. E.g, ``transforms.RandomCrop``
@@ -29,7 +30,7 @@ class SEMEION(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
         download: bool = True,
diff --git a/torchvision/datasets/stanford_cars.py b/torchvision/datasets/stanford_cars.py
index 5818648d81d..c029ed0d358 100644
--- a/torchvision/datasets/stanford_cars.py
+++ b/torchvision/datasets/stanford_cars.py
@@ -1,5 +1,5 @@
 import pathlib
-from typing import Any, Callable, Optional, Tuple
+from typing import Any, Callable, Optional, Tuple, Union
 
 from PIL import Image
 
@@ -21,7 +21,7 @@ class StanfordCars(VisionDataset):
         This class needs `scipy <https://docs.scipy.org/doc/>`_ to load target files from `.mat` format.
 
     Args:
-        root (string): Root directory of dataset
+        root (str or ``pathlib.Path``): Root directory of dataset
         split (string, optional): The dataset split, supports ``"train"`` (default) or ``"test"``.
         transform (callable, optional): A function/transform that takes in a PIL image
             and returns a transformed version. E.g, ``transforms.RandomCrop``
@@ -35,7 +35,7 @@ class StanfordCars(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, pathlib.Path],
         split: str = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/stl10.py b/torchvision/datasets/stl10.py
index fbfd6650a97..90ff41738eb 100644
--- a/torchvision/datasets/stl10.py
+++ b/torchvision/datasets/stl10.py
@@ -1,5 +1,6 @@
 import os.path
-from typing import Any, Callable, cast, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, cast, Optional, Tuple, Union
 
 import numpy as np
 from PIL import Image
@@ -12,7 +13,7 @@ class STL10(VisionDataset):
     """`STL10 <https://cs.stanford.edu/~acoates/stl10/>`_ Dataset.
 
     Args:
-        root (string): Root directory of dataset where directory
+        root (str or ``pathlib.Path``): Root directory of dataset where directory
             ``stl10_binary`` exists.
         split (string): One of {'train', 'test', 'unlabeled', 'train+unlabeled'}.
             Accordingly, dataset is selected.
@@ -45,7 +46,7 @@ class STL10(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         folds: Optional[int] = None,
         transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/sun397.py b/torchvision/datasets/sun397.py
index a32644b940c..4db0a3cf237 100644
--- a/torchvision/datasets/sun397.py
+++ b/torchvision/datasets/sun397.py
@@ -1,5 +1,5 @@
 from pathlib import Path
-from typing import Any, Callable, Optional, Tuple
+from typing import Any, Callable, Optional, Tuple, Union
 
 import PIL.Image
 
@@ -14,7 +14,7 @@ class SUN397(VisionDataset):
     397 categories with 108'754 images.
 
     Args:
-        root (string): Root directory of the dataset.
+        root (str or ``pathlib.Path``): Root directory of the dataset.
         transform (callable, optional): A function/transform that takes in a PIL image and returns a transformed
             version. E.g, ``transforms.RandomCrop``.
         target_transform (callable, optional): A function/transform that takes in the target and transforms it.
@@ -28,7 +28,7 @@ class SUN397(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
         download: bool = False,
diff --git a/torchvision/datasets/svhn.py b/torchvision/datasets/svhn.py
index ece384e2daa..5d20d7db7e3 100644
--- a/torchvision/datasets/svhn.py
+++ b/torchvision/datasets/svhn.py
@@ -1,5 +1,6 @@
 import os.path
-from typing import Any, Callable, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Optional, Tuple, Union
 
 import numpy as np
 from PIL import Image
@@ -19,7 +20,7 @@ class SVHN(VisionDataset):
         This class needs `scipy <https://docs.scipy.org/doc/>`_ to load data from `.mat` format.
 
     Args:
-        root (string): Root directory of the dataset where the data is stored.
+        root (str or ``pathlib.Path``): Root directory of the dataset where the data is stored.
         split (string): One of {'train', 'test', 'extra'}.
             Accordingly dataset is selected. 'extra' is Extra training set.
         transform (callable, optional): A function/transform that takes in a PIL image
@@ -52,7 +53,7 @@ class SVHN(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/ucf101.py b/torchvision/datasets/ucf101.py
index 7dece183060..935f8ad41c7 100644
--- a/torchvision/datasets/ucf101.py
+++ b/torchvision/datasets/ucf101.py
@@ -1,5 +1,6 @@
 import os
-from typing import Any, Callable, Dict, List, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 
 from torch import Tensor
 
@@ -28,7 +29,7 @@ class UCF101(VisionDataset):
     Internally, it uses a VideoClips object to handle clip creation.
 
     Args:
-        root (string): Root directory of the UCF101 Dataset.
+        root (str or ``pathlib.Path``): Root directory of the UCF101 Dataset.
         annotation_path (str): path to the folder containing the split files;
             see docstring above for download instructions of these files
         frames_per_clip (int): number of frames in a clip.
@@ -52,7 +53,7 @@ class UCF101(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         annotation_path: str,
         frames_per_clip: int,
         step_between_clips: int = 1,
diff --git a/torchvision/datasets/usps.py b/torchvision/datasets/usps.py
index 25f54d57cc9..9c681e79f6c 100644
--- a/torchvision/datasets/usps.py
+++ b/torchvision/datasets/usps.py
@@ -1,5 +1,6 @@
 import os
-from typing import Any, Callable, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, Optional, Tuple, Union
 
 import numpy as np
 from PIL import Image
@@ -15,7 +16,7 @@ class USPS(VisionDataset):
     and make pixel values in ``[0, 255]``.
 
     Args:
-        root (string): Root directory of dataset to store``USPS`` data files.
+        root (str or ``pathlib.Path``): Root directory of dataset to store``USPS`` data files.
         train (bool, optional): If True, creates dataset from ``usps.bz2``,
             otherwise from ``usps.t.bz2``.
         transform (callable, optional): A function/transform that takes in a PIL image
@@ -43,7 +44,7 @@ class USPS(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         train: bool = True,
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/vision.py b/torchvision/datasets/vision.py
index 82debfc881f..e524c67e263 100644
--- a/torchvision/datasets/vision.py
+++ b/torchvision/datasets/vision.py
@@ -1,5 +1,6 @@
 import os
-from typing import Any, Callable, List, Optional, Tuple
+from pathlib import Path
+from typing import Any, Callable, List, Optional, Tuple, Union
 
 import torch.utils.data as data
 
@@ -29,7 +30,7 @@ class VisionDataset(data.Dataset):
 
     def __init__(
         self,
-        root: str = None,  # type: ignore[assignment]
+        root: Union[str, Path] = None,  # type: ignore[assignment]
         transforms: Optional[Callable] = None,
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
diff --git a/torchvision/datasets/voc.py b/torchvision/datasets/voc.py
index cf76b8e5926..0f0e84c84fa 100644
--- a/torchvision/datasets/voc.py
+++ b/torchvision/datasets/voc.py
@@ -1,18 +1,18 @@
 import collections
 import os
+from pathlib import Path
+from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 from xml.etree.ElementTree import Element as ET_Element
 
-from .vision import VisionDataset
-
 try:
     from defusedxml.ElementTree import parse as ET_parse
 except ImportError:
     from xml.etree.ElementTree import parse as ET_parse
-from typing import Any, Callable, Dict, List, Optional, Tuple
 
 from PIL import Image
 
 from .utils import download_and_extract_archive, verify_str_arg
+from .vision import VisionDataset
 
 DATASET_YEAR_DICT = {
     "2012": {
@@ -67,7 +67,7 @@ class _VOCBase(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         year: str = "2012",
         image_set: str = "train",
         download: bool = False,
@@ -121,7 +121,7 @@ class VOCSegmentation(_VOCBase):
     """`Pascal VOC <http://host.robots.ox.ac.uk/pascal/VOC/>`_ Segmentation Dataset.
 
     Args:
-        root (string): Root directory of the VOC Dataset.
+        root (str or ``pathlib.Path``): Root directory of the VOC Dataset.
         year (string, optional): The dataset year, supports years ``"2007"`` to ``"2012"``.
         image_set (string, optional): Select the image_set to use, ``"train"``, ``"trainval"`` or ``"val"``. If
             ``year=="2007"``, can also be ``"test"``.
@@ -165,7 +165,7 @@ class VOCDetection(_VOCBase):
     """`Pascal VOC <http://host.robots.ox.ac.uk/pascal/VOC/>`_ Detection Dataset.
 
     Args:
-        root (string): Root directory of the VOC Dataset.
+        root (str or ``pathlib.Path``): Root directory of the VOC Dataset.
         year (string, optional): The dataset year, supports years ``"2007"`` to ``"2012"``.
         image_set (string, optional): Select the image_set to use, ``"train"``, ``"trainval"`` or ``"val"``. If
             ``year=="2007"``, can also be ``"test"``.
diff --git a/torchvision/datasets/widerface.py b/torchvision/datasets/widerface.py
index 020a5e5415c..90f80b7175b 100644
--- a/torchvision/datasets/widerface.py
+++ b/torchvision/datasets/widerface.py
@@ -1,5 +1,7 @@
 import os
 from os.path import abspath, expanduser
+from pathlib import Path
+
 from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 
 import torch
@@ -13,7 +15,7 @@ class WIDERFace(VisionDataset):
     """`WIDERFace <http://shuoyang1213.me/WIDERFACE/>`_ Dataset.
 
     Args:
-        root (string): Root directory where images and annotations are downloaded to.
+        root (str or ``pathlib.Path``): Root directory where images and annotations are downloaded to.
             Expects the following folder structure if download=False:
 
             .. code::
@@ -55,7 +57,7 @@ class WIDERFace(VisionDataset):
 
     def __init__(
         self,
-        root: str,
+        root: Union[str, Path],
         split: str = "train",
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,