Cleanup prototype datasets CI and related things (#6944)

* remove prototype datasets from CI

* move encoded features to prototype datasets namespace

* remove decoding transforms

* [REVERT ME] reinstate prototype datasets CI

* Revert "[REVERT ME] reinstate prototype datasets CI"

This reverts commit 215fb185cf6be5be7adf0388116c77acc9a5d3f3.

torchvision/prototype/datasets/utils/__init__.py

 from . import _internal  # usort: skip
 from ._dataset import Dataset
+from ._encoded import EncodedData, EncodedImage
 from ._resource import GDriveResource, HttpResource, KaggleDownloadResource, ManualDownloadResource, OnlineResource

torchvision/prototype/features/_encoded.py → torchvision/prototype/datasets/utils/_encoded.py

@@ -6,9 +6,9 @@ from typing import Any, BinaryIO, Optional, Tuple, Type, TypeVar, Union
 
 import PIL.Image
 import torch
-from torchvision.prototype.utils._internal import fromfile, ReadOnlyTensorBuffer
 
-from ._feature import _Feature
+from torchvision.prototype.features._feature import _Feature
+from torchvision.prototype.utils._internal import fromfile, ReadOnlyTensorBuffer
 
 D = TypeVar("D", bound="EncodedData")
 

torchvision/prototype/features/__init__.py

 from ._bounding_box import BoundingBox, BoundingBoxFormat
-from ._encoded import EncodedData, EncodedImage
 from ._feature import _Feature, FillType, FillTypeJIT, InputType, InputTypeJIT, is_simple_tensor
 from ._image import ColorSpace, Image, ImageType, ImageTypeJIT, TensorImageType, TensorImageTypeJIT
 from ._label import Label, OneHotLabel

torchvision/prototype/transforms/__init__.py

@@ -52,6 +52,6 @@ from ._misc import (
     TransposeDimensions,
 )
 from ._temporal import UniformTemporalSubsample
-from ._type_conversion import DecodeImage, LabelToOneHot, PILToTensor, ToImagePIL, ToImageTensor, ToPILImage
+from ._type_conversion import LabelToOneHot, PILToTensor, ToImagePIL, ToImageTensor, ToPILImage
 
 from ._deprecated import Grayscale, RandomGrayscale, ToTensor  # usort: skip

torchvision/prototype/transforms/_type_conversion.py

@@ -9,13 +9,6 @@ from torchvision.prototype import features
 from torchvision.prototype.transforms import functional as F, Transform
 
 
-class DecodeImage(Transform):
-    _transformed_types = (features.EncodedImage,)
-
-    def _transform(self, inpt: torch.Tensor, params: Dict[str, Any]) -> features.Image:
-        return F.decode_image_with_pil(inpt)  # type: ignore[no-any-return]
-
-
 class LabelToOneHot(Transform):
     _transformed_types = (features.Label,)
 

torchvision/prototype/transforms/functional/__init__.py

@@ -166,13 +166,6 @@ from ._misc import (
     normalize_video,
 )
 from ._temporal import uniform_temporal_subsample, uniform_temporal_subsample_video
-from ._type_conversion import (
-    decode_image_with_pil,
-    decode_video_with_av,
-    pil_to_tensor,
-    to_image_pil,
-    to_image_tensor,
-    to_pil_image,
-)
+from ._type_conversion import pil_to_tensor, to_image_pil, to_image_tensor, to_pil_image
 
 from ._deprecated import get_image_size, rgb_to_grayscale, to_grayscale, to_tensor  # usort: skip

torchvision/prototype/transforms/functional/_type_conversion.py

-from typing import Any, Dict, Tuple, Union
+from typing import Union
 
 import numpy as np
 import PIL.Image
 import torch
-from torchvision.io.video import read_video
 from torchvision.prototype import features
-from torchvision.prototype.utils._internal import ReadOnlyTensorBuffer
 from torchvision.transforms import functional as _F
 
 
-@torch.jit.unused
-def decode_image_with_pil(encoded_image: torch.Tensor) -> features.Image:
-    image = torch.as_tensor(np.array(PIL.Image.open(ReadOnlyTensorBuffer(encoded_image)), copy=True))
-    if image.ndim == 2:
-        image = image.unsqueeze(2)
-    return features.Image(image.permute(2, 0, 1))
-
-
-@torch.jit.unused
-def decode_video_with_av(encoded_video: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor, Dict[str, Any]]:
-    import unittest.mock
-
-    with unittest.mock.patch("torchvision.io.video.os.path.exists", return_value=True):
-        return read_video(ReadOnlyTensorBuffer(encoded_video))  # type: ignore[arg-type]
-
-
 @torch.jit.unused
 def to_image_tensor(image: Union[torch.Tensor, PIL.Image.Image, np.ndarray]) -> features.Image:
     if isinstance(image, np.ndarray):