pytorch/vision (torchvision)
### [`v0.19.0`](https://togithub.com/pytorch/vision/releases/tag/v0.19.0): Torchvision 0.19 release
[Compare Source](https://togithub.com/pytorch/vision/compare/v0.18.1...0.19.0)
### Highlights
##### Encoding / Decoding images
Torchvision is extending its encoding/decoding capabilities. For this version, **we added a GIF decoder** which is available as `torchvision.io.decode_gif(raw_tensor)`, `torchvision.io.decode_image(raw_tensor)`, and `torchvision.io.read_image(path_to_image)`.
We also **added support for jpeg GPU encoding** in `torchvision.io.encode_jpeg()`. This is 10X faster than the existing CPU jpeg encoder.
[Read more on the docs!](https://pytorch.org/vision/stable/io.html)
Stay tuned for more improvements coming in the next versions. We plan to improve jpeg GPU decoding, and add more image decoders (webp in particular).
##### Resizing according to the longest edge of an image
It is now possible to resize images by setting `torchvision.transforms.v2.Resize(max_size=N)`: this will resize the longest edge of the image exactly to `max_size`, making sure the image dimension don't exceed this value. [Read more on the docs!](https://pytorch.org/vision/stable/generated/torchvision.transforms.v2.Resize.html#torchvision.transforms.v2.Resize)
### Detailed changes
#### Bug Fixes
\[datasets] `SBDataset`: Only download noval file when image_set='train_noval' ([#8475](https://togithub.com/pytorch/vision/issues/8475))
\[datasets] Update the download url in class `EMNIST` ([#8350](https://togithub.com/pytorch/vision/issues/8350))
\[io] Fix compilation error when there is no `libjpeg` ([#8342](https://togithub.com/pytorch/vision/issues/8342))
\[reference scripts] Fix use of `cutmix_alpha` in classification training references ([#8448](https://togithub.com/pytorch/vision/issues/8448))
\[utils] Allow `K=1` in `draw_keypoints` ([#8439](https://togithub.com/pytorch/vision/issues/8439))
#### New Features
\[io] Add decoder for GIF images (`decode_gif()`, `decode_image()`,`read_image()`) ([#8406](https://togithub.com/pytorch/vision/issues/8406), [#8419](https://togithub.com/pytorch/vision/issues/8419))
\[transforms] Add `GaussianNoise` transform ([#8381](https://togithub.com/pytorch/vision/issues/8381))
#### Improvements
\[transforms] Allow v2 `Resize` to resize longer edge exactly to `max_size` ([#8459](https://togithub.com/pytorch/vision/issues/8459))
\[transforms] Add `min_area` parameter to `SanitizeBoundingBox` ([#7735](https://togithub.com/pytorch/vision/issues/7735))
\[transforms] Make `adjust_hue()` work with `numpy 2.0` ([#8463](https://togithub.com/pytorch/vision/issues/8463))
\[transforms] Enable one-hot-encoded labels in` MixUp ` and` CutMix ` ([#8427](https://togithub.com/pytorch/vision/issues/8427))
\[transforms] Create kernel on-device for `transforms.functional.gaussian_blur` ([#8426](https://togithub.com/pytorch/vision/issues/8426))
\[io] Adding GPU acceleration to `encode_jpeg` (10X faster than CPU encoder) ([#8391](https://togithub.com/pytorch/vision/issues/8391))
\[io] `read_video`: accept `BytesIO` objects on `pyav` backend ([#8442](https://togithub.com/pytorch/vision/issues/8442))
\[io] Add compatibility with FFMPEG 7.0 ([#8408](https://togithub.com/pytorch/vision/issues/8408))
\[datasets] Add extra to install `gdown` ([#8430](https://togithub.com/pytorch/vision/issues/8430))
\[datasets] Support encoded `RLE` format in for` COCO ` segmentations ([#8387](https://togithub.com/pytorch/vision/issues/8387))
\[datasets] Added binary cat vs dog classification target type to Oxford pet dataset ([#8388](https://togithub.com/pytorch/vision/issues/8388))
\[datasets] Return labels for `FER2013` if possible ([#8452](https://togithub.com/pytorch/vision/issues/8452))
\[ops] Force use of `torch.compile` on deterministic `roi_align` implementation ([#8436](https://togithub.com/pytorch/vision/issues/8436))
\[utils] add float support to `utils.draw_bounding_boxes()` ([#8328](https://togithub.com/pytorch/vision/issues/8328))
\[feature_extraction] Add concrete_args to feature extraction tracing. ([#8393](https://togithub.com/pytorch/vision/issues/8393))
\[Docs] Various documentation improvements ([#8429](https://togithub.com/pytorch/vision/issues/8429), [#8467](https://togithub.com/pytorch/vision/issues/8467), [#8469](https://togithub.com/pytorch/vision/issues/8469), [#8332](https://togithub.com/pytorch/vision/issues/8332), [#8262](https://togithub.com/pytorch/vision/issues/8262), [#8341](https://togithub.com/pytorch/vision/issues/8341), [#8392](https://togithub.com/pytorch/vision/issues/8392), [#8386](https://togithub.com/pytorch/vision/issues/8386), [#8385](https://togithub.com/pytorch/vision/issues/8385), [#8411](https://togithub.com/pytorch/vision/issues/8411)).
\[Tests] Various testing improvements ([#8454](https://togithub.com/pytorch/vision/issues/8454), [#8418](https://togithub.com/pytorch/vision/issues/8418), [#8480](https://togithub.com/pytorch/vision/issues/8480), [#8455](https://togithub.com/pytorch/vision/issues/8455))
\[Code quality] Various code quality improvements ([#8404](https://togithub.com/pytorch/vision/issues/8404), [#8402](https://togithub.com/pytorch/vision/issues/8402), [#8345](https://togithub.com/pytorch/vision/issues/8345), [#8335](https://togithub.com/pytorch/vision/issues/8335), [#8481](https://togithub.com/pytorch/vision/issues/8481), [#8334](https://togithub.com/pytorch/vision/issues/8334), [#8384](https://togithub.com/pytorch/vision/issues/8384), [#8451](https://togithub.com/pytorch/vision/issues/8451), [#8470](https://togithub.com/pytorch/vision/issues/8470), [#8413](https://togithub.com/pytorch/vision/issues/8413), [#8414](https://togithub.com/pytorch/vision/issues/8414), [#8416](https://togithub.com/pytorch/vision/issues/8416), [#8412](https://togithub.com/pytorch/vision/issues/8412))
#### Contributors
We're grateful for our community, which helps us improve torchvision by submitting issues and PRs, and providing feedback and suggestions. The following persons have contributed patches for this release:
Adam J. Stewart ahmadsharif1, AJS Payne, Andrew Lingg, Andrey Talman, Anner, Antoine Broyelle, cdzhan, deekay42, drhead, Edward Z. Yang, Emin Orhan, Fangjun Kuang, G, haarisr, Huy Do, Jack Newsom, JavaZero, Mahdi Lamb, Mantas, Nicolas Hug, Nicolas Hug , nihui, Richard Barnes , Richard Zou, Richie Bendall, Robert-André Mauchin, Ross Wightman, Siddarth Ijju, vfdev
Configuration
📅 Schedule: Branch creation - At any time (no schedule defined), Automerge - At any time (no schedule defined).
🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.
♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.
🔕 Ignore: Close this PR and you won't be reminded about this update again.
[ ] If you want to rebase/retry this PR, check this box
This PR contains the following updates:
^0.18.0->^0.19.0Release Notes
pytorch/vision (torchvision)
### [`v0.19.0`](https://togithub.com/pytorch/vision/releases/tag/v0.19.0): Torchvision 0.19 release [Compare Source](https://togithub.com/pytorch/vision/compare/v0.18.1...0.19.0) ### Highlights ##### Encoding / Decoding images Torchvision is extending its encoding/decoding capabilities. For this version, **we added a GIF decoder** which is available as `torchvision.io.decode_gif(raw_tensor)`, `torchvision.io.decode_image(raw_tensor)`, and `torchvision.io.read_image(path_to_image)`. We also **added support for jpeg GPU encoding** in `torchvision.io.encode_jpeg()`. This is 10X faster than the existing CPU jpeg encoder. [Read more on the docs!](https://pytorch.org/vision/stable/io.html) Stay tuned for more improvements coming in the next versions. We plan to improve jpeg GPU decoding, and add more image decoders (webp in particular). ##### Resizing according to the longest edge of an image It is now possible to resize images by setting `torchvision.transforms.v2.Resize(max_size=N)`: this will resize the longest edge of the image exactly to `max_size`, making sure the image dimension don't exceed this value. [Read more on the docs!](https://pytorch.org/vision/stable/generated/torchvision.transforms.v2.Resize.html#torchvision.transforms.v2.Resize) ### Detailed changes #### Bug Fixes \[datasets] `SBDataset`: Only download noval file when image_set='train_noval' ([#8475](https://togithub.com/pytorch/vision/issues/8475)) \[datasets] Update the download url in class `EMNIST` ([#8350](https://togithub.com/pytorch/vision/issues/8350)) \[io] Fix compilation error when there is no `libjpeg` ([#8342](https://togithub.com/pytorch/vision/issues/8342)) \[reference scripts] Fix use of `cutmix_alpha` in classification training references ([#8448](https://togithub.com/pytorch/vision/issues/8448)) \[utils] Allow `K=1` in `draw_keypoints` ([#8439](https://togithub.com/pytorch/vision/issues/8439)) #### New Features \[io] Add decoder for GIF images (`decode_gif()`, `decode_image()`,`read_image()`) ([#8406](https://togithub.com/pytorch/vision/issues/8406), [#8419](https://togithub.com/pytorch/vision/issues/8419)) \[transforms] Add `GaussianNoise` transform ([#8381](https://togithub.com/pytorch/vision/issues/8381)) #### Improvements \[transforms] Allow v2 `Resize` to resize longer edge exactly to `max_size` ([#8459](https://togithub.com/pytorch/vision/issues/8459)) \[transforms] Add `min_area` parameter to `SanitizeBoundingBox` ([#7735](https://togithub.com/pytorch/vision/issues/7735)) \[transforms] Make `adjust_hue()` work with `numpy 2.0` ([#8463](https://togithub.com/pytorch/vision/issues/8463)) \[transforms] Enable one-hot-encoded labels in` MixUp ` and` CutMix ` ([#8427](https://togithub.com/pytorch/vision/issues/8427)) \[transforms] Create kernel on-device for `transforms.functional.gaussian_blur` ([#8426](https://togithub.com/pytorch/vision/issues/8426)) \[io] Adding GPU acceleration to `encode_jpeg` (10X faster than CPU encoder) ([#8391](https://togithub.com/pytorch/vision/issues/8391)) \[io] `read_video`: accept `BytesIO` objects on `pyav` backend ([#8442](https://togithub.com/pytorch/vision/issues/8442)) \[io] Add compatibility with FFMPEG 7.0 ([#8408](https://togithub.com/pytorch/vision/issues/8408)) \[datasets] Add extra to install `gdown` ([#8430](https://togithub.com/pytorch/vision/issues/8430)) \[datasets] Support encoded `RLE` format in for` COCO ` segmentations ([#8387](https://togithub.com/pytorch/vision/issues/8387)) \[datasets] Added binary cat vs dog classification target type to Oxford pet dataset ([#8388](https://togithub.com/pytorch/vision/issues/8388)) \[datasets] Return labels for `FER2013` if possible ([#8452](https://togithub.com/pytorch/vision/issues/8452)) \[ops] Force use of `torch.compile` on deterministic `roi_align` implementation ([#8436](https://togithub.com/pytorch/vision/issues/8436)) \[utils] add float support to `utils.draw_bounding_boxes()` ([#8328](https://togithub.com/pytorch/vision/issues/8328)) \[feature_extraction] Add concrete_args to feature extraction tracing. ([#8393](https://togithub.com/pytorch/vision/issues/8393)) \[Docs] Various documentation improvements ([#8429](https://togithub.com/pytorch/vision/issues/8429), [#8467](https://togithub.com/pytorch/vision/issues/8467), [#8469](https://togithub.com/pytorch/vision/issues/8469), [#8332](https://togithub.com/pytorch/vision/issues/8332), [#8262](https://togithub.com/pytorch/vision/issues/8262), [#8341](https://togithub.com/pytorch/vision/issues/8341), [#8392](https://togithub.com/pytorch/vision/issues/8392), [#8386](https://togithub.com/pytorch/vision/issues/8386), [#8385](https://togithub.com/pytorch/vision/issues/8385), [#8411](https://togithub.com/pytorch/vision/issues/8411)). \[Tests] Various testing improvements ([#8454](https://togithub.com/pytorch/vision/issues/8454), [#8418](https://togithub.com/pytorch/vision/issues/8418), [#8480](https://togithub.com/pytorch/vision/issues/8480), [#8455](https://togithub.com/pytorch/vision/issues/8455)) \[Code quality] Various code quality improvements ([#8404](https://togithub.com/pytorch/vision/issues/8404), [#8402](https://togithub.com/pytorch/vision/issues/8402), [#8345](https://togithub.com/pytorch/vision/issues/8345), [#8335](https://togithub.com/pytorch/vision/issues/8335), [#8481](https://togithub.com/pytorch/vision/issues/8481), [#8334](https://togithub.com/pytorch/vision/issues/8334), [#8384](https://togithub.com/pytorch/vision/issues/8384), [#8451](https://togithub.com/pytorch/vision/issues/8451), [#8470](https://togithub.com/pytorch/vision/issues/8470), [#8413](https://togithub.com/pytorch/vision/issues/8413), [#8414](https://togithub.com/pytorch/vision/issues/8414), [#8416](https://togithub.com/pytorch/vision/issues/8416), [#8412](https://togithub.com/pytorch/vision/issues/8412)) #### Contributors We're grateful for our community, which helps us improve torchvision by submitting issues and PRs, and providing feedback and suggestions. The following persons have contributed patches for this release: Adam J. Stewart ahmadsharif1, AJS Payne, Andrew Lingg, Andrey Talman, Anner, Antoine Broyelle, cdzhan, deekay42, drhead, Edward Z. Yang, Emin Orhan, Fangjun Kuang, G, haarisr, Huy Do, Jack Newsom, JavaZero, Mahdi Lamb, Mantas, Nicolas Hug, Nicolas Hug , nihui, Richard Barnes , Richard Zou, Richie Bendall, Robert-André Mauchin, Ross Wightman, Siddarth Ijju, vfdevConfiguration
📅 Schedule: Branch creation - At any time (no schedule defined), Automerge - At any time (no schedule defined).
🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.
♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.
🔕 Ignore: Close this PR and you won't be reminded about this update again.
This PR was generated by Mend Renovate. View the repository job log.