[fbsync] fix antialias handling on resize and resized_crop (#6913)

Reviewed By: NicolasHug

Differential Revision: D41265189

fbshipit-source-id: cb611984d934c9fb1a534770bcadcefa753c5e19

Author: Yosua Michael Maranatha

torchvision/prototype/features/_bounding_box.py

@@ -78,7 +78,7 @@ def resize(  # type: ignore[override]
         size: List[int],
         interpolation: InterpolationMode = InterpolationMode.BILINEAR,
         max_size: Optional[int] = None,
-        antialias: bool = False,
+        antialias: Optional[bool] = None,
     ) -> BoundingBox:
         output, spatial_size = self._F.resize_bounding_box(
             self.as_subclass(torch.Tensor), spatial_size=self.spatial_size, size=size, max_size=max_size
@@ -105,7 +105,7 @@ def resized_crop(
         width: int,
         size: List[int],
         interpolation: InterpolationMode = InterpolationMode.BILINEAR,
-        antialias: bool = False,
+        antialias: Optional[bool] = None,
     ) -> BoundingBox:
         output, spatial_size = self._F.resized_crop_bounding_box(
             self.as_subclass(torch.Tensor), self.format, top, left, height, width, size=size

torchvision/prototype/features/_feature.py

@@ -164,7 +164,7 @@ def resize(  # type: ignore[override]
         size: List[int],
         interpolation: InterpolationMode = InterpolationMode.BILINEAR,
         max_size: Optional[int] = None,
-        antialias: bool = False,
+        antialias: Optional[bool] = None,
     ) -> _Feature:
         return self
 
@@ -182,7 +182,7 @@ def resized_crop(
         width: int,
         size: List[int],
         interpolation: InterpolationMode = InterpolationMode.BILINEAR,
-        antialias: bool = False,
+        antialias: Optional[bool] = None,
     ) -> _Feature:
         return self
 

torchvision/prototype/features/_image.py

@@ -123,7 +123,7 @@ def resize(  # type: ignore[override]
         size: List[int],
         interpolation: InterpolationMode = InterpolationMode.BILINEAR,
         max_size: Optional[int] = None,
-        antialias: bool = False,
+        antialias: Optional[bool] = None,
     ) -> Image:
         output = self._F.resize_image_tensor(
             self.as_subclass(torch.Tensor), size, interpolation=interpolation, max_size=max_size, antialias=antialias
@@ -146,7 +146,7 @@ def resized_crop(
         width: int,
         size: List[int],
         interpolation: InterpolationMode = InterpolationMode.BILINEAR,
-        antialias: bool = False,
+        antialias: Optional[bool] = None,
     ) -> Image:
         output = self._F.resized_crop_image_tensor(
             self.as_subclass(torch.Tensor),

torchvision/prototype/features/_mask.py

@@ -49,7 +49,7 @@ def resize(  # type: ignore[override]
         size: List[int],
         interpolation: InterpolationMode = InterpolationMode.NEAREST,
         max_size: Optional[int] = None,
-        antialias: bool = False,
+        antialias: Optional[bool] = None,
     ) -> Mask:
         output = self._F.resize_mask(self.as_subclass(torch.Tensor), size, max_size=max_size)
         return Mask.wrap_like(self, output)
@@ -70,7 +70,7 @@ def resized_crop(
         width: int,
         size: List[int],
         interpolation: InterpolationMode = InterpolationMode.NEAREST,
-        antialias: bool = False,
+        antialias: Optional[bool] = None,
     ) -> Mask:
         output = self._F.resized_crop_mask(self.as_subclass(torch.Tensor), top, left, height, width, size=size)
         return Mask.wrap_like(self, output)

torchvision/prototype/features/_video.py

@@ -79,7 +79,7 @@ def resize(  # type: ignore[override]
         size: List[int],
         interpolation: InterpolationMode = InterpolationMode.BILINEAR,
         max_size: Optional[int] = None,
-        antialias: bool = False,
+        antialias: Optional[bool] = None,
     ) -> Video:
         output = self._F.resize_video(
             self.as_subclass(torch.Tensor),
@@ -106,7 +106,7 @@ def resized_crop(
         width: int,
         size: List[int],
         interpolation: InterpolationMode = InterpolationMode.BILINEAR,
-        antialias: bool = False,
+        antialias: Optional[bool] = None,
     ) -> Video:
         output = self._F.resized_crop_video(
             self.as_subclass(torch.Tensor),

torchvision/prototype/transforms/functional/_geometry.py

@@ -115,8 +115,9 @@ def resize_image_tensor(
     size: List[int],
     interpolation: InterpolationMode = InterpolationMode.BILINEAR,
     max_size: Optional[int] = None,
-    antialias: bool = False,
+    antialias: Optional[bool] = None,
 ) -> torch.Tensor:
+    antialias = False if antialias is None else antialias
     align_corners: Optional[bool] = None
     if interpolation == InterpolationMode.BILINEAR or interpolation == InterpolationMode.BICUBIC:
         align_corners = False
@@ -196,7 +197,7 @@ def resize_video(
     size: List[int],
     interpolation: InterpolationMode = InterpolationMode.BILINEAR,
     max_size: Optional[int] = None,
-    antialias: bool = False,
+    antialias: Optional[bool] = None,
 ) -> torch.Tensor:
     return resize_image_tensor(video, size=size, interpolation=interpolation, max_size=max_size, antialias=antialias)
 
@@ -209,10 +210,8 @@ def resize(
     antialias: Optional[bool] = None,
 ) -> features.InputTypeJIT:
     if isinstance(inpt, torch.Tensor) and (torch.jit.is_scripting() or not isinstance(inpt, features._Feature)):
-        antialias = False if antialias is None else antialias
         return resize_image_tensor(inpt, size, interpolation=interpolation, max_size=max_size, antialias=antialias)
     elif isinstance(inpt, features._Feature):
-        antialias = False if antialias is None else antialias
         return inpt.resize(size, interpolation=interpolation, max_size=max_size, antialias=antialias)
     else:
         if antialias is not None and not antialias:
@@ -1396,7 +1395,7 @@ def resized_crop_image_tensor(
     width: int,
     size: List[int],
     interpolation: InterpolationMode = InterpolationMode.BILINEAR,
-    antialias: bool = False,
+    antialias: Optional[bool] = None,
 ) -> torch.Tensor:
     image = crop_image_tensor(image, top, left, height, width)
     return resize_image_tensor(image, size, interpolation=interpolation, antialias=antialias)
@@ -1449,7 +1448,7 @@ def resized_crop_video(
     width: int,
     size: List[int],
     interpolation: InterpolationMode = InterpolationMode.BILINEAR,
-    antialias: bool = False,
+    antialias: Optional[bool] = None,
 ) -> torch.Tensor:
     return resized_crop_image_tensor(
         video, top, left, height, width, antialias=antialias, size=size, interpolation=interpolation
@@ -1467,12 +1466,10 @@ def resized_crop(
     antialias: Optional[bool] = None,
 ) -> features.InputTypeJIT:
     if isinstance(inpt, torch.Tensor) and (torch.jit.is_scripting() or not isinstance(inpt, features._Feature)):
-        antialias = False if antialias is None else antialias
         return resized_crop_image_tensor(
             inpt, top, left, height, width, antialias=antialias, size=size, interpolation=interpolation
         )
     elif isinstance(inpt, features._Feature):
-        antialias = False if antialias is None else antialias
         return inpt.resized_crop(top, left, height, width, antialias=antialias, size=size, interpolation=interpolation)
     else:
         return resized_crop_image_pil(inpt, top, left, height, width, size=size, interpolation=interpolation)