cj-mills/hagrid-sample-30k-384p

This dataset contains 31,833 images from HaGRID (HAnd Gesture Recognition Image Dataset) downscaled to 384p. The original dataset is 716GB and contains 552,992 1080p images. I created this sample for a tutorial so readers can use the dataset in the free tiers of Google Colab and Kaggle Notebooks.

Original Authors:

• Alexander Kapitanov
• Andrey Makhlyarchuk
• Karina Kvanchiani

Original Dataset Links

• GitHub
• Kaggle Datasets Page

Object Classes

['call',
 'no_gesture',
 'dislike',
 'fist',
 'four',
 'like',
 'mute',
 'ok',
 'one',
 'palm',
 'peace',
 'peace_inverted',
 'rock',
 'stop',
 'stop_inverted',
 'three',
 'three2',
 'two_up',
 'two_up_inverted']

Annotations

• bboxes: [top-left-X-position, top-left-Y-position, width, height]
• Multiply top-left-X-position and width values by the image width and multiply top-left-Y-position and height values by the image height.

  	00005c9c-3548-4a8f-9d0b-2dd4aff37fc9
  bboxes	[[0.23925175, 0.28595301, 0.25055143, 0.20777627]]
  labels	[call]
  leading_hand	right
  leading_conf	1
  user_id	5a389ffe1bed6660a59f4586c7d8fe2770785e5bf79b09334aa951f6f119c024