Rename `convert_bounding_box_format` => `convert_format_bounding_box` (#6582)

* Rename `convert_bounding_box_format` => `convert_format_bounding_box`

* Add missed replacement.

test/test_prototype_transforms_functional.py

@@ -19,7 +19,7 @@ from prototype_common_utils import (
 from torch import jit
 from torchvision.prototype import features
 from torchvision.prototype.transforms.functional._geometry import _center_crop_compute_padding
-from torchvision.prototype.transforms.functional._meta import convert_bounding_box_format
+from torchvision.prototype.transforms.functional._meta import convert_format_bounding_box
 from torchvision.transforms.functional import _get_perspective_coeffs
 
 
@@ -633,7 +633,7 @@ def test_correctness_affine_bounding_box(angle, translate, scale, shear, center)
         affine_matrix = _compute_affine_matrix(angle_, translate_, scale_, shear_, center_)
         affine_matrix = affine_matrix[:2, :]
 
-        bbox_xyxy = convert_bounding_box_format(
+        bbox_xyxy = convert_format_bounding_box(
             bbox, old_format=bbox.format, new_format=features.BoundingBoxFormat.XYXY
         )
         points = np.array(
@@ -658,7 +658,7 @@ def test_correctness_affine_bounding_box(angle, translate, scale, shear, center)
             dtype=torch.float32,
             device=bbox.device,
         )
-        return convert_bounding_box_format(
+        return convert_format_bounding_box(
             out_bbox, old_format=features.BoundingBoxFormat.XYXY, new_format=bbox.format, copy=False
         )
 
@@ -835,7 +835,7 @@ def test_correctness_rotate_bounding_box(angle, expand, center):
         affine_matrix = affine_matrix[:2, :]
 
         image_size = bbox.image_size
-        bbox_xyxy = convert_bounding_box_format(
+        bbox_xyxy = convert_format_bounding_box(
             bbox, old_format=bbox.format, new_format=features.BoundingBoxFormat.XYXY
         )
         points = np.array(
@@ -876,7 +876,7 @@ def test_correctness_rotate_bounding_box(angle, expand, center):
             dtype=torch.float32,
             device=bbox.device,
         )
-        return convert_bounding_box_format(
+        return convert_format_bounding_box(
             out_bbox, old_format=features.BoundingBoxFormat.XYXY, new_format=bbox.format, copy=False
         )
 
@@ -1097,7 +1097,7 @@ def test_correctness_crop_bounding_box(device, format, top, left, height, width,
     ]
     in_boxes = features.BoundingBox(in_boxes, format=features.BoundingBoxFormat.XYXY, image_size=size, device=device)
     if format != features.BoundingBoxFormat.XYXY:
-        in_boxes = convert_bounding_box_format(in_boxes, features.BoundingBoxFormat.XYXY, format)
+        in_boxes = convert_format_bounding_box(in_boxes, features.BoundingBoxFormat.XYXY, format)
 
     output_boxes = F.crop_bounding_box(
         in_boxes,
@@ -1107,7 +1107,7 @@ def test_correctness_crop_bounding_box(device, format, top, left, height, width,
     )
 
     if format != features.BoundingBoxFormat.XYXY:
-        output_boxes = convert_bounding_box_format(output_boxes, format, features.BoundingBoxFormat.XYXY)
+        output_boxes = convert_format_bounding_box(output_boxes, format, features.BoundingBoxFormat.XYXY)
 
     torch.testing.assert_close(output_boxes.tolist(), expected_bboxes)
 
@@ -1213,12 +1213,12 @@ def test_correctness_resized_crop_bounding_box(device, format, top, left, height
         in_boxes, format=features.BoundingBoxFormat.XYXY, image_size=image_size, device=device
     )
     if format != features.BoundingBoxFormat.XYXY:
-        in_boxes = convert_bounding_box_format(in_boxes, features.BoundingBoxFormat.XYXY, format)
+        in_boxes = convert_format_bounding_box(in_boxes, features.BoundingBoxFormat.XYXY, format)
 
     output_boxes = F.resized_crop_bounding_box(in_boxes, format, top, left, height, width, size)
 
     if format != features.BoundingBoxFormat.XYXY:
-        output_boxes = convert_bounding_box_format(output_boxes, format, features.BoundingBoxFormat.XYXY)
+        output_boxes = convert_format_bounding_box(output_boxes, format, features.BoundingBoxFormat.XYXY)
 
     torch.testing.assert_close(output_boxes, expected_bboxes)
 
@@ -1268,12 +1268,12 @@ def test_correctness_pad_bounding_box(device, padding):
 
         bbox_format = bbox.format
         bbox_dtype = bbox.dtype
-        bbox = convert_bounding_box_format(bbox, old_format=bbox_format, new_format=features.BoundingBoxFormat.XYXY)
+        bbox = convert_format_bounding_box(bbox, old_format=bbox_format, new_format=features.BoundingBoxFormat.XYXY)
 
         bbox[0::2] += pad_left
         bbox[1::2] += pad_up
 
-        bbox = convert_bounding_box_format(
+        bbox = convert_format_bounding_box(
             bbox, old_format=features.BoundingBoxFormat.XYXY, new_format=bbox_format, copy=False
         )
         if bbox.dtype != bbox_dtype:
@@ -1396,7 +1396,7 @@ def test_correctness_perspective_bounding_box(device, startpoints, endpoints):
             ]
         )
 
-        bbox_xyxy = convert_bounding_box_format(
+        bbox_xyxy = convert_format_bounding_box(
             bbox, old_format=bbox.format, new_format=features.BoundingBoxFormat.XYXY
         )
         points = np.array(
@@ -1423,7 +1423,7 @@ def test_correctness_perspective_bounding_box(device, startpoints, endpoints):
             dtype=torch.float32,
             device=bbox.device,
         )
-        return convert_bounding_box_format(
+        return convert_format_bounding_box(
             out_bbox, old_format=features.BoundingBoxFormat.XYXY, new_format=bbox.format, copy=False
         )
 
@@ -1528,7 +1528,7 @@ def test_correctness_center_crop_bounding_box(device, output_size):
     def _compute_expected_bbox(bbox, output_size_):
         format_ = bbox.format
         image_size_ = bbox.image_size
-        bbox = convert_bounding_box_format(bbox, format_, features.BoundingBoxFormat.XYWH)
+        bbox = convert_format_bounding_box(bbox, format_, features.BoundingBoxFormat.XYWH)
 
         if len(output_size_) == 1:
             output_size_.append(output_size_[-1])
@@ -1548,7 +1548,7 @@ def test_correctness_center_crop_bounding_box(device, output_size):
             dtype=bbox.dtype,
             device=bbox.device,
         )
-        return convert_bounding_box_format(out_bbox, features.BoundingBoxFormat.XYWH, format_, copy=False)
+        return convert_format_bounding_box(out_bbox, features.BoundingBoxFormat.XYWH, format_, copy=False)
 
     for bboxes in make_bounding_boxes(
         image_sizes=[(32, 32), (24, 33), (32, 25)],

torchvision/prototype/features/_bounding_box.py

@@ -65,7 +65,7 @@ class BoundingBox(_Feature):
             format = BoundingBoxFormat.from_str(format.upper())
 
         return BoundingBox.new_like(
-            self, self._F.convert_bounding_box_format(self, old_format=self.format, new_format=format), format=format
+            self, self._F.convert_format_bounding_box(self, old_format=self.format, new_format=format), format=format
         )
 
     def horizontal_flip(self) -> BoundingBox:

torchvision/prototype/transforms/_augment.py

@@ -254,7 +254,7 @@ class SimpleCopyPaste(_RandomApplyTransform):
         # There is a similar +1 in other reference implementations:
         # https://github.com/pytorch/vision/blob/b6feccbc4387766b76a3e22b13815dbbbfa87c0f/torchvision/models/detection/roi_heads.py#L418-L422
         xyxy_boxes[:, 2:] += 1
-        boxes = F.convert_bounding_box_format(
+        boxes = F.convert_format_bounding_box(
             xyxy_boxes, old_format=features.BoundingBoxFormat.XYXY, new_format=bbox_format, copy=False
         )
         out_target["boxes"] = torch.cat([boxes, paste_boxes])
@@ -263,7 +263,7 @@ class SimpleCopyPaste(_RandomApplyTransform):
         out_target["labels"] = torch.cat([labels, paste_labels])
 
         # Check for degenerated boxes and remove them
-        boxes = F.convert_bounding_box_format(
+        boxes = F.convert_format_bounding_box(
             out_target["boxes"], old_format=bbox_format, new_format=features.BoundingBoxFormat.XYXY
         )
         degenerate_boxes = boxes[:, 2:] <= boxes[:, :2]

torchvision/prototype/transforms/_geometry.py

@@ -655,7 +655,7 @@ class RandomIoUCrop(Transform):
                     continue
 
                 # check for any valid boxes with centers within the crop area
-                xyxy_bboxes = F.convert_bounding_box_format(
+                xyxy_bboxes = F.convert_format_bounding_box(
                     bboxes, old_format=bboxes.format, new_format=features.BoundingBoxFormat.XYXY, copy=True
                 )
                 cx = 0.5 * (xyxy_bboxes[..., 0] + xyxy_bboxes[..., 2])

torchvision/prototype/transforms/_meta.py

@@ -17,7 +17,7 @@ class ConvertBoundingBoxFormat(Transform):
         self.format = format
 
     def _transform(self, inpt: features.BoundingBox, params: Dict[str, Any]) -> features.BoundingBox:
-        output = F.convert_bounding_box_format(inpt, old_format=inpt.format, new_format=params["format"])
+        output = F.convert_format_bounding_box(inpt, old_format=inpt.format, new_format=params["format"])
         return features.BoundingBox.new_like(inpt, output, format=params["format"])
 
 

torchvision/prototype/transforms/_misc.py

@@ -163,7 +163,7 @@ class RemoveSmallBoundingBoxes(Transform):
         #  be in XYXY format only to calculate the width and height internally. Thus, if the box is in XYWH or CXCYWH
         #  format,we need to convert first just to afterwards compute the width and height again, although they were
         #  there in the first place for these formats.
-        bounding_box = F.convert_bounding_box_format(
+        bounding_box = F.convert_format_bounding_box(
             bounding_box, old_format=bounding_box.format, new_format=features.BoundingBoxFormat.XYXY
         )
         valid_indices = remove_small_boxes(bounding_box, min_size=self.min_size)

torchvision/prototype/transforms/functional/__init__.py

@@ -3,7 +3,7 @@
 from torchvision.transforms import InterpolationMode  # usort: skip
 from ._meta import (
     clamp_bounding_box,
-    convert_bounding_box_format,
+    convert_format_bounding_box,
     convert_color_space_image_tensor,
     convert_color_space_image_pil,
     convert_color_space,

torchvision/prototype/transforms/functional/_geometry.py

@@ -17,7 +17,7 @@ from torchvision.transforms.functional import (
 )
 from torchvision.transforms.functional_tensor import _parse_pad_padding
 
-from ._meta import convert_bounding_box_format, get_dimensions_image_pil, get_dimensions_image_tensor
+from ._meta import convert_format_bounding_box, get_dimensions_image_pil, get_dimensions_image_tensor
 
 
 # shortcut type
@@ -37,13 +37,13 @@ def horizontal_flip_bounding_box(
 ) -> torch.Tensor:
     shape = bounding_box.shape
 
-    bounding_box = convert_bounding_box_format(
+    bounding_box = convert_format_bounding_box(
         bounding_box, old_format=format, new_format=features.BoundingBoxFormat.XYXY
     ).view(-1, 4)
 
     bounding_box[:, [0, 2]] = image_size[1] - bounding_box[:, [2, 0]]
 
-    return convert_bounding_box_format(
+    return convert_format_bounding_box(
         bounding_box, old_format=features.BoundingBoxFormat.XYXY, new_format=format, copy=False
     ).view(shape)
 
@@ -70,13 +70,13 @@ def vertical_flip_bounding_box(
 ) -> torch.Tensor:
     shape = bounding_box.shape
 
-    bounding_box = convert_bounding_box_format(
+    bounding_box = convert_format_bounding_box(
         bounding_box, old_format=format, new_format=features.BoundingBoxFormat.XYXY
     ).view(-1, 4)
 
     bounding_box[:, [1, 3]] = image_size[0] - bounding_box[:, [3, 1]]
 
-    return convert_bounding_box_format(
+    return convert_format_bounding_box(
         bounding_box, old_format=features.BoundingBoxFormat.XYXY, new_format=format, copy=False
     ).view(shape)
 
@@ -362,7 +362,7 @@ def affine_bounding_box(
     center: Optional[List[float]] = None,
 ) -> torch.Tensor:
     original_shape = bounding_box.shape
-    bounding_box = convert_bounding_box_format(
+    bounding_box = convert_format_bounding_box(
         bounding_box, old_format=format, new_format=features.BoundingBoxFormat.XYXY
     ).view(-1, 4)
 
@@ -370,7 +370,7 @@ def affine_bounding_box(
 
     # out_bboxes should be of shape [N boxes, 4]
 
-    return convert_bounding_box_format(
+    return convert_format_bounding_box(
         out_bboxes, old_format=features.BoundingBoxFormat.XYXY, new_format=format, copy=False
     ).view(original_shape)
 
@@ -530,13 +530,13 @@ def rotate_bounding_box(
         center = None
 
     original_shape = bounding_box.shape
-    bounding_box = convert_bounding_box_format(
+    bounding_box = convert_format_bounding_box(
         bounding_box, old_format=format, new_format=features.BoundingBoxFormat.XYXY
     ).view(-1, 4)
 
     out_bboxes = _affine_bounding_box_xyxy(bounding_box, image_size, angle=-angle, center=center, expand=expand)
 
-    return convert_bounding_box_format(
+    return convert_format_bounding_box(
         out_bboxes, old_format=features.BoundingBoxFormat.XYXY, new_format=format, copy=False
     ).view(original_shape)
 
@@ -704,7 +704,7 @@ def crop_bounding_box(
     top: int,
     left: int,
 ) -> torch.Tensor:
-    bounding_box = convert_bounding_box_format(
+    bounding_box = convert_format_bounding_box(
         bounding_box, old_format=format, new_format=features.BoundingBoxFormat.XYXY
     )
 
@@ -712,7 +712,7 @@ def crop_bounding_box(
     bounding_box[..., 0::2] -= left
     bounding_box[..., 1::2] -= top
 
-    return convert_bounding_box_format(
+    return convert_format_bounding_box(
         bounding_box, old_format=features.BoundingBoxFormat.XYXY, new_format=format, copy=False
     )
 
@@ -758,7 +758,7 @@ def perspective_bounding_box(
         raise ValueError("Argument perspective_coeffs should have 8 float values")
 
     original_shape = bounding_box.shape
-    bounding_box = convert_bounding_box_format(
+    bounding_box = convert_format_bounding_box(
         bounding_box, old_format=format, new_format=features.BoundingBoxFormat.XYXY
     ).view(-1, 4)
 
@@ -828,7 +828,7 @@ def perspective_bounding_box(
 
     # out_bboxes should be of shape [N boxes, 4]
 
-    return convert_bounding_box_format(
+    return convert_format_bounding_box(
         out_bboxes, old_format=features.BoundingBoxFormat.XYXY, new_format=format, copy=False
     ).view(original_shape)
 
@@ -900,7 +900,7 @@ def elastic_bounding_box(
     displacement = displacement.to(bounding_box.device)
 
     original_shape = bounding_box.shape
-    bounding_box = convert_bounding_box_format(
+    bounding_box = convert_format_bounding_box(
         bounding_box, old_format=format, new_format=features.BoundingBoxFormat.XYXY
     ).view(-1, 4)
 
@@ -926,7 +926,7 @@ def elastic_bounding_box(
     out_bbox_maxs, _ = torch.max(transformed_points, dim=1)
     out_bboxes = torch.cat([out_bbox_mins, out_bbox_maxs], dim=1)
 
-    return convert_bounding_box_format(
+    return convert_format_bounding_box(
         out_bboxes, old_format=features.BoundingBoxFormat.XYXY, new_format=format, copy=False
     ).view(original_shape)
 

torchvision/prototype/transforms/functional/_meta.py

@@ -79,7 +79,7 @@ def _xyxy_to_cxcywh(xyxy: torch.Tensor) -> torch.Tensor:
     return torch.stack((cx, cy, w, h), dim=-1)
 
 
-def convert_bounding_box_format(
+def convert_format_bounding_box(
     bounding_box: torch.Tensor, old_format: BoundingBoxFormat, new_format: BoundingBoxFormat, copy: bool = True
 ) -> torch.Tensor:
     if new_format == old_format:
@@ -106,10 +106,10 @@ def clamp_bounding_box(
 ) -> torch.Tensor:
     # TODO: (PERF) Possible speed up clamping if we have different implementations for each bbox format.
     # Not sure if they yield equivalent results.
-    xyxy_boxes = convert_bounding_box_format(bounding_box, format, BoundingBoxFormat.XYXY)
+    xyxy_boxes = convert_format_bounding_box(bounding_box, format, BoundingBoxFormat.XYXY)
     xyxy_boxes[..., 0::2].clamp_(min=0, max=image_size[1])
     xyxy_boxes[..., 1::2].clamp_(min=0, max=image_size[0])
-    return convert_bounding_box_format(xyxy_boxes, BoundingBoxFormat.XYXY, format, copy=False)
+    return convert_format_bounding_box(xyxy_boxes, BoundingBoxFormat.XYXY, format, copy=False)
 
 
 def _split_alpha(image: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor]: