-
transform = T.Compose(
[
T.Resize(min_size, max_size),
T.RandomHorizontalFlip(flip_prob),
T.ToTensor(),
normalize_transform,
]
th…
-
Are there any plans to allow for custom data augmentation (scale, flip, rotate, etc.) through TuriCreate for the `turicreate.image_classifier`?
If not, are there any recommended python libraries fo…
-
Hello,
In the dataset_format documentation, it is specified that "The different input channels MUST have the same geometry (same shape, spacing (if applicable) etc.) and must be co-registered (if …
-
### Describe the bug
keepdim is supposed to retain the dimensions of the input tensor passed. However here passing a CHW tensor into an augmentation results in a batch dim being added.
### Repr…
-
## Reference
- [【OCR技术系列之三】大批量生成文字训练集](https://www.cnblogs.com/skyfsm/p/8436820.html)
## Brief
- 文字扭曲
- 背景噪声(椒盐)
- 文字位置(设置文字的中心点)
- 笔画粘连(膨胀来模拟)
- 笔画断裂(腐蚀来模拟)
- 文字倾斜(文字旋转)
- 多种字体
-
Hello Pytorch vision people !
I am currently working on a project that requires lots of image augmentations
to perform better. And I believe this is not only my case. When reading
about …
-
Hi,
In the training code, I noticed `random_flip` being used as an augmentation.
Did you see any impact of this in terms of spatial captions?
e.g. if the caption says `a toy to the left of cat` an…
-
Hello, thank you for your outstanding work. I would like to know if you have compared the fda method and the random image augmentation method, because from the visual results, there are some similarit…
-
Hi madebyollin!
Thank you so much for your awesome work and kind reply.
Sorry to bother you again. While trying to finetune taesd decoder, I found that the images are getting **a little dim** and …
-
### Search before asking
- [X] I have searched the YOLOv8 [issues](https://github.com/ultralytics/ultralytics/issues) and [discussions](https://github.com/ultralytics/ultralytics/discussions) and fou…