Google Summer of Code - 2019

Certificate

Organization : TensorFlow - [TensorFlow Datasets]

Project Link : https://summerofcode.withgoogle.com/archive/2019/projects/6537043688030208/

Mentors : Marcin Michalski, Etienne Pot, Pierre Ruyssen

Pull Requests

• Link to all my PRs : https://github.com/tensorflow/datasets/pulls/ChanchalKumarMaji
  • #596 : Add tests for COCO dataset.
  • #613 : Add Downsampled ImageNet dataset.
  • #628 : Add missing files to setup.py.
  • #635 : Multi-Names support for ClassLabel.
  • #651 : Improve usability of batch_size in as_dataset function.
  • #670 : Add visualization.py to tfds.
  • #701 : Read directly from archives on datasets with many records.
  • #716 : Add COCO_2017 and COCO_2017_Panoptic datasets.
  • #743 : Use pytest-xdist to speed up kokoro tests.
  • #874 : Launch S3 to structured datasets.
  • #891 : Add missing citations.
  • #839 : Add cifar-10.1 dataset.
  • #924 : Launch S3 to video datasets.

Reviews

• Link to some of the PRs which I reviewed
  • #263 : Add mini_imagenet dataset.
  • #311 : Add tiered_imagenet dataset.
  • #359 : Add AFLW2000-3D dataset.
  • #361 : Add Quickdraw Sketch RNN dataset.
  • #436 : Add Cartoon Set dataset.
  • #439 : Add DiscoFuse dataset.
  • #562 : Add CMATERdb dataset.
  • #582 : Add yelp_polarity_review dataset.
  • #835 : Add Fruits360 dataset.
  • #888 : Fix broken link on README.md.
  • #928 : Add tiny_imagenet dataset.

Future Work

• #677 : Sorted TFrecords writer for beam.
• I will continue my work with the TensorFlow team.