Unified input for resized crop op #2396
Conversation
| top (int): Vertical component of the top left corner of the crop box. | ||
| left (int): Horizontal component of the top left corner of the crop box. | ||
| height (int): Height of the crop box. | ||
| width (int): Width of the crop box. | ||
| size (sequence or int): Desired output size. Same semantics as ``resize``. | ||
| interpolation (int, optional): Desired interpolation. Default is | ||
| ``PIL.Image.BILINEAR``. | ||
| interpolation (int, optional): Desired interpolation. Default is ``PIL.Image.BILINEAR``. |
There was a problem hiding this comment.
Technically, this is not "optional" (Optional is for None).
There was a problem hiding this comment.
yeah, this is a potential issue with the documentation: it has a default value, so the user doesn't need to specify it, so it is in some sense optional, but I see your point.
There was a problem hiding this comment.
Seems like it is not the only usage like that. Maybe, we can leave it for now ?
Codecov Report
@@ Coverage Diff @@
## master #2396 +/- ##
==========================================
+ Coverage 70.65% 70.68% +0.02%
==========================================
Files 94 94
Lines 7897 7905 +8
Branches 1241 1245 +4
==========================================
+ Hits 5580 5588 +8
+ Misses 1934 1932 -2
- Partials 383 385 +2
Continue to review full report at Codecov.
|
| @@ -532,7 +532,7 @@ def resize(img: Tensor, size: List[int], interpolation: int = 2) -> Tensor: | |||
| elif len(size) < 2: | |||
| size_w, size_h = size[0], size[0] | |||
| else: | |||
| size_w, size_h = size[0], size[1] | |||
| size_w, size_h = size[1], size[0] # Convention (h, w) | |||
torchvision/transforms/transforms.py
Outdated
| if 0 < w <= width and 0 < h <= height: | ||
| i = random.randint(0, height - h) | ||
| j = random.randint(0, width - w) | ||
| if 0 < w < width and 0 < h < height: |
There was a problem hiding this comment.
was the <= a bug before? Or is ti because random.randint is inclusive on end value and torch.randint is exclusive on end value?
torchvision/transforms/transforms.py
Outdated
| i = torch.randint(0, height - h, size=(1,)).item() | ||
| j = torch.randint(0, width - w, size=(1,)).item() | ||
| if 0 < w <= width and 0 < h <= height: | ||
| i = torch.randint(0, max(height - h, 1), size=(1,)).item() |
There was a problem hiding this comment.
don't we need to add a +1 to account for the previous behavior, instead of making it max(..., 1)?
Related to #2292
Description:
T.ResizedCrop,F.resized_cropwont pass due to Unified input for resize op #2394)Blocked by #2394