Skip to content

Releases: ashnair1/COCO-Assistant

0.4.0

23 Aug 11:48
Compare
Choose a tag to compare

Changes:

  • Drop Python 3.6 support
  • Use zillionare's cookiecutter-pypackage as project template
  • Improved documentation
  • Standardise docstring syles (#32)

Bugfixes:

v0.3.5

06 May 16:40
Compare
Choose a tag to compare

Changes:

  • Replace os with pathlib for managing paths.
  • Fixed bug in det2seg where small annotations were covered by larger annotations.
  • Faster category removal by @xychen9459 (#43).

Minor release for testing CircleCI deployment

14 Mar 18:18
Compare
Choose a tag to compare

Changes:

  • Add support for merging datasets with different categories
  • Adding colour palette to segmasks is now optional
  • Ignore hidden files

v0.3.1

18 May 06:54
c6ffa1b
Compare
Choose a tag to compare

Bugfix for coco_stats by @Lplenka (#15)

v0.3.0

19 Apr 19:48
Compare
Choose a tag to compare

Changes introduced:

  1. combine has been renamed to merge
  2. Support for merging annotations only.

Deprecated features:

Post v0.3.0, converter functionality will not be supported. This was motivated by two reasons:

  1. There are too many conversion formats. Trying to include even the most popular ones makes the project unwieldy. Better to use one of the readily available scripts online.
  2. Until now, the repository only supported TFRecord. Tensorflow is a large library and it does not make sense for it to be a project requirement when it's only used for a singular task of converting to TFRecord.

v0.2.0

28 Nov 07:54
Compare
Choose a tag to compare

New Features:

  1. Generate anchors from the dataset using K-means.
  2. Generate segmentation masks from dataset.

Modifications:

  1. Modified category removal to accept a list of categories as opposed to entering categories one after the other.

v0.1.0

07 Oct 19:10
Compare
Choose a tag to compare

Helper for dealing with MS COCO annotations. Currently supports the following functionalities:

  1. Merge datasets.
  2. Remove specific category from dataset.
  3. Generate annotations statistics - distribution of object areas and category distribution
  4. Annotation visualiser for viewing the entire dataset.
  5. Converter for converting annotations to TFRecord format