Add Tensor support for some transforms #1104
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Codecov Report
@@ Coverage Diff @@
## master #1104 +/- ##
==========================================
+ Coverage 63.92% 64.89% +0.96%
==========================================
Files 68 68
Lines 5406 5438 +32
Branches 829 843 +14
==========================================
+ Hits 3456 3529 +73
+ Misses 1707 1650 -57
- Partials 243 259 +16
Continue to review full report at Codecov.
|
fmassa
changed the title
soumith
approved these changes
Jul 26, 2019
| @@ -234,26 +243,42 @@ def resize(img, size, interpolation=Image.BILINEAR): | |||
| Returns: | |||
| PIL Image: Resized image. | |||
| """ | |||
| if not _is_pil_image(img): | |||
| if not (_is_pil_image(img) or isinstance(img, torch.Tensor)): | |||
There was a problem hiding this comment.
the comment above needs to be updated that it takes torch.Tensor (and has to specify what range the Tensor's values have to be)
| @@ -362,16 +387,19 @@ def crop(img, i, j, h, w): | |||
| Returns: | |||
| PIL Image: Cropped image. | |||
| """ | |||
| if not _is_pil_image(img): | |||
| if not (_is_pil_image(img) or isinstance(img, torch.Tensor)): | |||
| Image.NEAREST: "nearest", | ||
| Image.BILINEAR: "bilinear" | ||
| } | ||
| should_unsqueeze = False |
There was a problem hiding this comment.
this is actually should_squeeze as you squeeze below in 278
| @@ -410,10 +438,13 @@ def hflip(img): | |||
| Returns: | |||
| PIL Image: Horizontall flipped image. | |||
| """ | |||
| if not _is_pil_image(img): | |||
| if not (_is_pil_image(img) or isinstance(img, torch.Tensor)): | |||
There was a problem hiding this comment.
same documentation comment as above
| raise TypeError('img should be PIL Image. Got {}'.format(type(img))) | ||
|
|
||
| return img.transpose(Image.FLIP_LEFT_RIGHT) | ||
| if _is_pil_image(img): |
There was a problem hiding this comment.
it's much cleaner to write explicit if/else, rather than have if shortcut to a return
| @@ -468,10 +499,13 @@ def vflip(img): | |||
| Returns: | |||
| PIL Image: Vertically flipped image. | |||
| """ | |||
| if not _is_pil_image(img): | |||
| if not (_is_pil_image(img) or isinstance(img, torch.Tensor)): | |||
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is a POC implementation for adding support for tensors in some of the transforms.
It will be specially useful for video, and also as a first step towards enabling (some of) the transforms to be traceable.