Computer vision

ENV

Python 3.9
jina 2.5.0

Describe the bug

If i try to dump an image blob to a io.bytesio object an error is thrown

from jina import Document
import io
d =  Document(uri='steam_data/image_store/8c/5b/8c5b265b9c533636.png')
output = io.BytesIO() 

(
    d
    .load_uri_to_image_blob()
    .dump_ima

Initially brought in pytorch/vision#4472 (review)

TestBoxArea, TestBoxIoU and TestGenBoxIoU are all very related and perform a single test (or two if we consider the newly added JIT test).
We can probably refactor the implementations so that they follow the same base class, and use pytest.parameterize, so that each configuration is displayed a

At the moment, from API there are two useful columns about tasks:

However, in tabs (views) there are possible columns tasks:created_at a

🚀 Feature

As reported by deepsource in here we abuse from using built-in input function in our functionality.

Motivation

We target to have a clean and healthy source code free of risk.

Pitch

Replace variable names whether it makes sense e.g. for image based functionality input -> image ; in l

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

Perhaps I missed something but it took me a while to realise that there was an __call__ on this object. Perhaps extending the example would help prevent people from having this issue in future.

https://github.com/scikit-image/scikit-image/blob/a655681010caff396a2f584686c524246ae359d3/skimage/transform/_geometric.py#L602

🚨🚨 Feature Request

• Related to an existing Issue
• A new implementation (Improvement, Extension)

If your feature will improve HUB

Need a way to check if a dataset already exists.

hub.empty throws an error if a dataset exists and hub.load throws an error if the dataset does not exist.

Need a way to check if a dataset already exists without throwing a

Could FeatureTools be implemented as an automated preprocessor to Autogluon, adding the ability to handle multi-entity problems (i.e. Data split across multiple normalised database tables)? So if you supply Autogluon with a list of Dataframes instead of a single Dataframe it would first invoke FeatureTools:

• take the multiple Dataframes (entities) and try to auto-infer the relationship betwee

WHAT

Make plot_duplicates output clearer.

WHY

The graph plotted by plot_duplicates output can be cluttered, especially when there are several duplicates found for an image. The titles of each image run into each other, which makes it hard to read the plot. Additionally, the titles can be hard to read, possibly due to a static dpi. An example can be seen below:

![imgdedup_plot_iss

Well #77 didn't work for me while resuming from checkpoint_18.pth. The problem is when we resume, the model and optimizer passed in the restore_from function are suitable for epoch less than 10 (till backbone is not training) because the cfg.TRAIN.START_EPOCH is 0 (passed in build_opt_lr function just before restore_from) initially so this mismatches the optimizer after backbone start training. So