{
  "id": 408144,
  "title": "Developing a Mobile App for Tunisian Sign Language Translation: Data Extraction and Model Training Challenges",
  "url": "/competitions/asl-signs/discussion/408144",
  "author_name": "",
  "post_date": "2023-05-09T17:01:29.330096500Z",
  "votes": 4,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hello there, Kaggle community.<br>\nThis competition inspired me to work on an academic project to create a mobile&nbsp;app that can translate Tunisian sign language.  I discovered a data source from which I extracted the landmarks and stored them in a parquet files&nbsp;in a manner similar to this competition data. However, I'm experiencing poor results&nbsp;and would like to learn from your answers.  I'm also struggling with how to implement the model for real-time prediction in the mobile app.<br>\nThis is the link for the <a href=\"https://www.kaggle.com/datasets/fouedayedi/tunisian-sign-language-landmarks\" target=\"_blank\">data</a>. And my <a href=\"https://www.kaggle.com/code/fouedayedi/lstm-baseline-prediction/notebook\" target=\"_blank\">starter notebook</a>.<br>\nThank you for any assistance you may offer.</p>",
  "messages": [
    {
      "id": "2251858",
      "postDate": "05/09/2023 17:01:29",
      "content": "<p>Hello there, Kaggle community.<br>\nThis competition inspired me to work on an academic project to create a mobile&nbsp;app that can translate Tunisian sign language.  I discovered a data source from which I extracted the landmarks and stored them in a parquet files&nbsp;in a manner similar to this competition data. However, I'm experiencing poor results&nbsp;and would like to learn from your answers.  I'm also struggling with how to implement the model for real-time prediction in the mobile app.<br>\nThis is the link for the <a href=\"https://www.kaggle.com/datasets/fouedayedi/tunisian-sign-language-landmarks\" target=\"_blank\">data</a>. And my <a href=\"https://www.kaggle.com/code/fouedayedi/lstm-baseline-prediction/notebook\" target=\"_blank\">starter notebook</a>.<br>\nThank you for any assistance you may offer.</p>",
      "rawMarkdown": "Hello there, Kaggle community.\nThis competition inspired me to work on an academic project to create a mobile app that can translate Tunisian sign language.  I discovered a data source from which I extracted the landmarks and stored them in a parquet files in a manner similar to this competition data. However, I'm experiencing poor results and would like to learn from your answers.  I'm also struggling with how to implement the model for real-time prediction in the mobile app.\nThis is the link for the [data](https://www.kaggle.com/datasets/fouedayedi/tunisian-sign-language-landmarks). And my [starter notebook](https://www.kaggle.com/code/fouedayedi/lstm-baseline-prediction/notebook).\nThank you for any assistance you may offer.",
      "votes": null
    },
    {
      "id": "2251864",
      "postDate": "05/09/2023 17:09:08",
      "content": "<p>That is a really cool project!</p>",
      "rawMarkdown": "That is a really cool project!",
      "votes": null
    },
    {
      "id": "2253108",
      "postDate": "05/09/2023 23:25:14",
      "content": "<p>Thank you for your kind words! <br>\nI appreciate your encouragement on my project. I'm looking forward to sharing the final results.</p>",
      "rawMarkdown": "Thank you for your kind words! \nI appreciate your encouragement on my project. I'm looking forward to sharing the final results.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2251864,
      "author_name": "sohier",
      "author_url": "",
      "post_date": "05/09/2023 17:09:08",
      "content": "<p>That is a really cool project!</p>",
      "votes": null,
      "replies": [
        {
          "id": 2253108,
          "author_name": "fouedayedi",
          "author_url": "",
          "post_date": "05/09/2023 23:25:14",
          "content": "<p>Thank you for your kind words! <br>\nI appreciate your encouragement on my project. I'm looking forward to sharing the final results.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2251858": "Hello there, Kaggle community.\nThis competition inspired me to work on an academic project to create a mobile app that can translate Tunisian sign language.  I discovered a data source from which I extracted the landmarks and stored them in a parquet files in a manner similar to this competition data. However, I'm experiencing poor results and would like to learn from your answers.  I'm also struggling with how to implement the model for real-time prediction in the mobile app.\nThis is the link for the [data](https://www.kaggle.com/datasets/fouedayedi/tunisian-sign-language-landmarks). And my [starter notebook](https://www.kaggle.com/code/fouedayedi/lstm-baseline-prediction/notebook).\nThank you for any assistance you may offer.",
    "2251864": "That is a really cool project!",
    "2253108": "Thank you for your kind words! \nI appreciate your encouragement on my project. I'm looking forward to sharing the final results."
  },
  "source": "meta"
}