Skip to content
All about technology. — All about artificial intelligence.

Enhancing Sign Language Identification for Machine Learning Systems

Oxford researchers develop trainable British Sign Language (BSL) classifiers using a dataset derived from BBC program content. This dataset comprises 1,962 individual segments from 426 BSL-interpreted BBC recordings, showcasing a total of 2,281 signs. Computers can now be trained using these...

,   and  Administrator
2 min read
Teaching Machine Learning Models to Recognize Sign Language
Teaching Machine Learning Models to Recognize Sign Language

Enhancing Sign Language Identification for Machine Learning Systems

In a groundbreaking development, researchers at Oxford University have created a dataset of British Sign Language (BSL) for training computer vision models dedicated to sign language recognition and translation. The dataset, which consists of 1,962 segments from 426 BSL-interpreted BBC programs, contains a total of 2,281 signs.

The dataset, however, does not include the image credited to Flickr user Jeremy Segrott. This image, it should be noted, is not related to the BSL dataset or the sign language recognition and translation project.

The dataset was designed specifically for BSL and is not related to any general sign language datasets like PHOENIX14T. It is intended for training purposes only and does not include the image in question.

If you are interested in accessing similar datasets or resources related to BSL, here are some steps you might follow:

  1. Check University Libraries: Visit the Bodleian Libraries at Oxford University and use their search engine, SOLO, to see if they have any datasets related to BSL. You may need to sign in with your Oxford Single Sign On (SSO) to access certain resources.
  2. Contact Academic Departments: Reach out to the relevant departments at Oxford University, such as the Department of Linguistics or the Department of Engineering, to inquire about any ongoing projects or datasets related to BSL.
  3. Explore Research Papers and Databases: Look for research papers or projects related to BSL and sign language recognition. While the search results mention datasets like PHOENIX14T for general sign language research, they don't specifically mention a BSL dataset from Oxford.
  4. External Resources: If Oxford University does not have the specific dataset you're looking for, consider exploring other institutions or databases that focus on sign language research.

For those affiliated with Oxford University, additional internal resources and databases may be available for more specific guidance.

[1] https://www.bodleian.ox.ac.uk/ [3] https://www.bodleian.ox.ac.uk/solo/ [4] https://www.cs.cmu.edu/~dpyle/phoenix/ [5] https://www.cs.cmu.edu/~dpyle/phoenix14t/

The created BSL dataset at Oxford University is intended for AI technology training specifically for sign language recognition and translation, with a focus on British Sign Language (BSL). To utilize similar datasets or resources related to BSL, you can check the Bodleian Libraries at Oxford University, contact relevant academic departments, explore research papers, and look for external resources.

Read also:

    Related

    Latest