6 datasets found
  1. h

    hagrid-sample-250k-384p

    • huggingface.co
    Updated Jul 3, 2023
    Share
    FacebookFacebook
    TwitterTwitter
    Email
    Click to copy link
    Link copied
    Close
    Cite
    Christian Mills (2023). hagrid-sample-250k-384p [Dataset]. https://huggingface.co/datasets/cj-mills/hagrid-sample-250k-384p
    Explore at:
    CroissantCroissant is a format for machine-learning datasets. Learn more about this at mlcommons.org/croissant.
    Dataset updated
    Jul 3, 2023
    Authors
    Christian Mills
    License

    Attribution-ShareAlike 4.0 (CC BY-SA 4.0)https://creativecommons.org/licenses/by-sa/4.0/
    License information was derived automatically

    Description

    This dataset contains 254,661 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'… See the full description on the dataset page: https://huggingface.co/datasets/cj-mills/hagrid-sample-250k-384p.

  2. h

    hagrid-sample-500k-384p

    • huggingface.co
    Updated Jul 3, 2023
    Share
    FacebookFacebook
    TwitterTwitter
    Email
    Click to copy link
    Link copied
    Close
    Cite
    Christian Mills (2023). hagrid-sample-500k-384p [Dataset]. https://huggingface.co/datasets/cj-mills/hagrid-sample-500k-384p
    Explore at:
    CroissantCroissant is a format for machine-learning datasets. Learn more about this at mlcommons.org/croissant.
    Dataset updated
    Jul 3, 2023
    Authors
    Christian Mills
    License

    Attribution-ShareAlike 4.0 (CC BY-SA 4.0)https://creativecommons.org/licenses/by-sa/4.0/
    License information was derived automatically

    Description

    This dataset contains 509,323 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'… See the full description on the dataset page: https://huggingface.co/datasets/cj-mills/hagrid-sample-500k-384p.

  3. h

    hagrid-sample-30k-384p

    • huggingface.co
    Updated Jul 2, 2023
    Share
    FacebookFacebook
    TwitterTwitter
    Email
    Click to copy link
    Link copied
    Close
    Cite
    Christian Mills (2023). hagrid-sample-30k-384p [Dataset]. https://huggingface.co/datasets/cj-mills/hagrid-sample-30k-384p
    Explore at:
    CroissantCroissant is a format for machine-learning datasets. Learn more about this at mlcommons.org/croissant.
    Dataset updated
    Jul 2, 2023
    Authors
    Christian Mills
    License

    Attribution-ShareAlike 4.0 (CC BY-SA 4.0)https://creativecommons.org/licenses/by-sa/4.0/
    License information was derived automatically

    Description

    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'… See the full description on the dataset page: https://huggingface.co/datasets/cj-mills/hagrid-sample-30k-384p.

  4. h

    hagrid-sample-120k-384p

    • huggingface.co
    Updated Jul 3, 2023
    Share
    FacebookFacebook
    TwitterTwitter
    Email
    Click to copy link
    Link copied
    Close
    Cite
    Christian Mills (2023). hagrid-sample-120k-384p [Dataset]. https://huggingface.co/datasets/cj-mills/hagrid-sample-120k-384p
    Explore at:
    CroissantCroissant is a format for machine-learning datasets. Learn more about this at mlcommons.org/croissant.
    Dataset updated
    Jul 3, 2023
    Authors
    Christian Mills
    License

    Attribution-ShareAlike 4.0 (CC BY-SA 4.0)https://creativecommons.org/licenses/by-sa/4.0/
    License information was derived automatically

    Description

    This dataset contains 127,331 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'… See the full description on the dataset page: https://huggingface.co/datasets/cj-mills/hagrid-sample-120k-384p.

  5. HaGRID Sample 120k 384p

    • kaggle.com
    Updated Sep 8, 2022
    Share
    FacebookFacebook
    TwitterTwitter
    Email
    Click to copy link
    Link copied
    Close
    Cite
    Innominate817 (2022). HaGRID Sample 120k 384p [Dataset]. https://www.kaggle.com/datasets/innominate817/hagrid-sample-120k-384p
    Explore at:
    CroissantCroissant is a format for machine-learning datasets. Learn more about this at mlcommons.org/croissant.
    Dataset updated
    Sep 8, 2022
    Dataset provided by
    Kagglehttp://kaggle.com/
    Authors
    Innominate817
    License

    Attribution-ShareAlike 4.0 (CC BY-SA 4.0)https://creativecommons.org/licenses/by-sa/4.0/
    License information was derived automatically

    Description

    This dataset contains 127,331 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:

    Original Dataset Links

    Kaggle Notebooks

    Initial IceVision YOLOX Tiny Results:

    • 30K 384p: 76.2509% COCOMetric after 20 epochs
    • 60K 384p: 78.4270% COCOMetric after 20 epochs
    • 120K 384p: 81.3696% COCOMetric after 20 epochs
    • 500K 384p: 81.5283% COCOMetric after 10 epochs

    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_handright
      leading_conf1
      user_id5a389ffe1bed6660a59f4586c7d8fe2770785e5bf79b09334aa951f6f119c024
  6. HaGRID Sample 30k 384p

    • kaggle.com
    Updated Sep 8, 2022
    Share
    FacebookFacebook
    TwitterTwitter
    Email
    Click to copy link
    Link copied
    Close
    Cite
    Innominate817 (2022). HaGRID Sample 30k 384p [Dataset]. https://www.kaggle.com/datasets/innominate817/hagrid-sample-30k-384p
    Explore at:
    CroissantCroissant is a format for machine-learning datasets. Learn more about this at mlcommons.org/croissant.
    Dataset updated
    Sep 8, 2022
    Dataset provided by
    Kagglehttp://kaggle.com/
    Authors
    Innominate817
    License

    Attribution-ShareAlike 4.0 (CC BY-SA 4.0)https://creativecommons.org/licenses/by-sa/4.0/
    License information was derived automatically

    Description

    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:

    Original Dataset Links

    Kaggle Notebooks

    • Training:
      • The free GPU tier for Kaggle Notebooks takes around 15 minutes per epoch.
      • A screen recording walking through the setup steps is available on Youtube (link). Timestamps are in the video description.
    • Inference

    Initial IceVision YOLOX Tiny Results:

    • 30K 384p: 76.2509% COCOMetric after 20 epochs
    • 60K 384p: 78.4270% COCOMetric after 20 epochs
    • 120K 384p: 81.3696% COCOMetric after 20 epochs
    • 500K 384p: 81.5283% COCOMetric after 10 epochs

    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_handright
      leading_conf1
      user_id5a389ffe1bed6660a59f4586c7d8fe2770785e5bf79b09334aa951f6f119c024
  7. Not seeing a result you expected?
    Learn how you can add new datasets to our index.

Share
FacebookFacebook
TwitterTwitter
Email
Click to copy link
Link copied
Close
Cite
Christian Mills (2023). hagrid-sample-250k-384p [Dataset]. https://huggingface.co/datasets/cj-mills/hagrid-sample-250k-384p

hagrid-sample-250k-384p

HaGRID Sample 250k 384p

cj-mills/hagrid-sample-250k-384p

Explore at:
CroissantCroissant is a format for machine-learning datasets. Learn more about this at mlcommons.org/croissant.
Dataset updated
Jul 3, 2023
Authors
Christian Mills
License

Attribution-ShareAlike 4.0 (CC BY-SA 4.0)https://creativecommons.org/licenses/by-sa/4.0/
License information was derived automatically

Description

This dataset contains 254,661 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'… See the full description on the dataset page: https://huggingface.co/datasets/cj-mills/hagrid-sample-250k-384p.

Search
Clear search
Close search
Google apps
Main menu