image-classification

Enhancement

A discussion in #614 revealed a good place for improvement - we should ensure that input image is continuous upon start of the augmentation pipeline. This could be implemented by adding image = np.ascontiguousarray(image) to image and mask targets.

A proposed place to add this call - somewhere at the beginning of A.Compose.__call__.

Describe the bug
When exporting a brush annotation as a PNG, the output is not mapped by the background colors specified in (Settings > Labeling Interface). In addition, when exporting as a JSON, the background colors for the attributes are not specified anywhere, leaving the values that were selected in the interface as arbitrary and as not linked to any of the outputs.

To Reproduce

My actions before raising this issue

• Read/searched the docs
• Searched past issues

Expected Behaviour

I want to access the traefik dashboard. I have cvat setup on a server with https. I modified the docker-compose.https.yml according to the comment for using the traefik dashboard. So included the lin

Hi,
I need to download the something-to-something and jester datasets. But the 20bn website "https://20bn.com" are not available for weeks, the error message is "503 Service Temporarily Unavailable".

I have already downloaded the video data of something-to-something v2, and I need the label dataset. For the Jester, I need both video and label data. Can someone share me the

Add PECOS model to TabularPredictor.

If you'd like to work on this, please respond to this GitHub issue.

It is recommended to follow the custom model tutorial for implementing the model into AutoGluon.

Currently the X argument of CleanLearning.fit() does not seem to support non-array data.
Perhaps this is due to the sklearn function check_X_y() called inside CleanLearning, which we could replace.
Or perhaps it's due to how the cross-validation is currently being implemented.

However these are both easy to improve to rid the restriction that only array data are supported.
Seems e