Maxim Zhiltsov

Results 94 issues of Maxim Zhiltsov

An ImageNet dataset can have any of the following layouts: ``` bash # Single subset, no relative paths /n234543/image1.jpg /n343523/image2.jpg # Multiple subsets, no relative paths /train/n345345/image1.jpg /test/n456456/image2.jpg # Single...

data formats

As such plugins are just python code, they can be specifically crafted by an attacker to be run by a user on opening a project. The possible solution is disallowing...

security

These things are not necessarily harmful in a CLI tool use cases, and they can even be desired in such scenario. However, they can be undesirable in a server-side use...

DOC
security

Currently, Datumaro only works with static datasets on the disk. However, it can be useful to provide an ability to input dataset from the standard input, or devices (IP cameras...

ENHANCE
cli

Currently, `datum filter` works in XML format and uses XPath for queries. JSON output can be more desired in some cases, can be combined with `jq`, for example. There are...

ENHANCE

Add checks for original datasets (probably, on PR to `master`) that they can be downloaded and imported.

data formats

When publishing a dataset and when training a model (to avoid bias) it could be useful to remove sensitive information like car plate numbers, names and other personal info on...

ENHANCE
good first issue

Review and update `extractor.py` and other files. > Changed in version 21.1.0: init=False injects \_\_attrs_init\_\_ Should fix our class patching in Bbox.

After releasing a dataset, it is often needs to be updated and maintained. In this process, or even in the process of experimenting and building the first version, it is...

ENHANCE

https://packaging.python.org/guides/publishing-package-distribution-releases-using-github-actions-ci-cd-workflows/ Should be executed on PRs to `master`.